UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The role of metamorphic and geochemical factors in the formation of gem corundum, spinel, and haüyne… Belley, Philippe Maxime 2019

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

Item Metadata

Download

Media
24-ubc_2019_may_belley_philippe.pdf [ 11.14MB ]
Metadata
JSON: 24-1.0378174.json
JSON-LD: 24-1.0378174-ld.json
RDF/XML (Pretty): 24-1.0378174-rdf.xml
RDF/JSON: 24-1.0378174-rdf.json
Turtle: 24-1.0378174-turtle.txt
N-Triples: 24-1.0378174-rdf-ntriples.txt
Original Record: 24-1.0378174-source.json
Full Text
24-1.0378174-fulltext.txt
Citation
24-1.0378174.ris

Full Text

  THE ROLE OF METAMORPHIC AND GEOCHEMICAL FACTORS IN THE FORMATION OF GEM CORUNDUM, SPINEL, AND HAÜYNE IN METACARBONATES OF THE LAKE HARBOUR GROUP, BAFFIN ISLAND, CANADA      by  PHILIPPE MAXIME BELLEY  B.Sc., University of Ottawa, 2014        A DISSERTATION SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in  THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES  (Geological Sciences)     THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)     April 2019     © Philippe Maxime Belley, 2019ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled:  The role of metamorphic and geochemical factors in the formation of gem corundum, spinel, and haüyne in metacarbonates of the Lake Harbour Group, Baffin Island, Canada  Submitted by Philippe Maxime Belley in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Geological Sciences.  Examining Committee: Dr. Lee A. Groat Supervisor  Dr. James K. Mortensen Supervisory Committee Member  Dr. Lori Kennedy Supervisory Committee Member  Dr. Tom Troczynski University Examiner  Dr. Maya Kopylova University Examiner  Dr. Ian Graham  External Examiner University of New South Wales  iii  Abstract The Lake Harbour Group (LHG) metacarbonates on Baffin Island contain occurrences of the gemstones sapphire (corundum), spinel (including cobalt-blue), and lapis lazuli (blue haüyne-rich rock). This dissertation uncovers the regional geologic processes (e.g., metamorphic history, metasomatism, protolith geochemistry) that influence gemstone potential by developing genetic models for the LHG gem mineral occurrences. Both barren and gem-bearing metacarbonates were studied using field examination and sampling, petrological (optical) and scanning electron microscope petrography, and whole rock geochemistry. Boron isotope geochemistry, thermodynamic modelling, and age-dating (zircon U-Pb and mica 40Ar-39Ar) were employed for selected occurrences. Corundum formation was made possible by three equally important sequential metamorphic reactions: (1) formation of nepheline, diopside, and K-feldspar (inferred) at granulite facies peak metamorphic conditions; (2) partial retrograde replacement of the peak assemblage by phlogopite, oligoclase, calcite, and scapolite; and (3) retrograde break-down of scapolite + nepheline to form albite, muscovite, corundum, and calcite. The corundum-forming reaction only occurs in a <100 °C window. Spinel and haüyne formed at granulite facies peak metamorphic conditions, and spinel remained stable through upper amphibolite facies retrogression. The corundum, spinel, and haüyne occurrences are interpreted to have different sedimentary protoliths with the exception of a few metasomatic spinel occurrences. The protoliths of the occurrences are interpreted to be: [A] impure dolomitic limestone (spinel); [B] dolomitic marl (spinel, corundum); [C] magnesite-rich evaporitic marl (spinel); and [D] evaporite (halite and anhydrite)-bearing dolomite-rich marl (haüyne). The meta-marls have Al/Si iv  abundances comparable to that expected in a siliciclastic mud. Calc-silicate rocks barren of gem minerals have lower Al/Si (sandier protoliths). Spinel-bearing calc-silicates have higher Al/Si relative to that of sapphire- and haüyne-bearing rocks. High abundances of Al to Si are crucial to spinel formation in Mg-rich calc-silicates. Spinel formation in impure dolomitic marbles requires low K activity and does not require high Al/Si. Cobalt enrichment at two spinel occurrences is localized and interpreted to represent geochemical features of the protolith. The genetic models provide broad gemstone exploration criteria for carbonate-bearing metasedimentary sequences such as the LHG.   v  Lay Summary Marble and associated rocks are major producers of commercial gemstones around the world. Such rocks occur on southern Baffin Island, and contain gem sapphire, spinel, and lapis lazuli. This thesis uses a collection of analytical methods to identify the large- and small-scale geological factors that result in gemstone formation. Factors leading to gem formation include changes in pressure and temperature conditions during metamorphism and having favourable chemical compositions in the initial sedimentary rocks. Sapphire formed via a three-step metamorphic process, most likely to occur near a regional fault, which transformed a magnesium-rich mixed mud and limestone rock into a sapphire-bearing metamorphic rock. Spinel formed from the metamorphism of impure magnesium-rich limestones and mixed mud-dolomite rock, where spinel formation is dependent on low potassium concentrations and high aluminium relative to silicon. Lapis lazuli formed from a mixed mud-carbonate rock containing evaporite salt impurities. These research results have helped develop exploration criteria.vi  Preface This dissertation is an original independent work by the author, Philippe Belley. All discussions and conclusions represent my own independent work. Portions of the thesis have been published or submitted as manuscripts in peer-reviewed journals. Much of the data presented in Chapter 3 was collected by previous workers and re-interpreted in the present dissertation. Co-authors provided several figures, as described below.  Chapters 1 and 2  Portions of Chapter 1 and the totality of Chapter 2 appear in the papers listed below in Chapters 3 and 4.  Chapter 3 Chapter 3 was published in the 2017 Gem Materials special issue of The Canadian Mineralogist, volume 55, pages 669-699 in the paper entitled “Origin of scapolite-hosted sapphire (corundum) near Kimmirut, Baffin Island, Nunavut, Canada by Philippe M. Belley (present author), T.J. Dzikowski, A. Fagan, J. Cempirek, L.A. Groat, J.K. Mortensen, M. Fayek, G. Giuliani, A.E. Fallick, & P. Gertzbein. T.J. Dzikowski contributed some petrographic observations (petrography was largely re-done independently by the present author) and collected electron microprobe (EPMA) data. A. Fagan (in his capacity with True North Gems Inc.) provided the geological map of the TNG Inc. property near Kimmirut. J. Cempirek collected EPMA data on, and re-calculated the formula of oxy-dravite. J.K. Mortensen conducted TIMS analysis of zircon. M. Fayek conducted the B isotope analysis of oxy-dravite using SIMS. The oxygen isotopic composition of corundum was measured by A.E. Fallick. Mica Ar-Ar data vii  was collected by Andrea Cade. The discussion is the complete independent work of the present author (Philippe Belley). Andrew Fagan provided a useful suggestion on Al-silicate/corundum stability that improved the discussion. The Al-Mg-Ca ternary plot was provided by G. Giuliani. The REE and oxygen isotope plots are after Dzikowski (2013).  Chapter 4 Chapter 4 has been accepted with revision in The Canadian Mineralogist. The paper, by Philippe M. Belley and Lee A. Groat, is entitled “Metacarbonate-hosted spinel on Baffin Island, Nunavut, Canada: Insights into the origin of gem spinel and cobalt-blue spinel.” Chapter 4 is the complete independent work of the present author (Philippe Belley). Lee Groat provided useful comments which improved the quality of Chapter 4 and the manuscript.  Chapter 5  Some of the new contributions on lapis lazuli were included in the spinel paper (Ch. 4), where lapis lazuli was compared to other metacarbonates.  Chapter 6 An adapted version of Chapter 6 will be submitted as a manuscript to a peer-reviewed journal. viii  Table of Contents  Abstract .......................................................................................................................................... iii Lay Summary .................................................................................................................................. v Preface............................................................................................................................................ vi Table of Contents ......................................................................................................................... viii List of Tables ................................................................................................................................ xii List of Figures .............................................................................................................................. xvi Acknowledgments...................................................................................................................... xxiv Chapter 1. Introduction and Literature Review .............................................................................. 1 1.1 Introduction ........................................................................................................................ 1 1.2 Geology of study area ........................................................................................................ 3 1.2.1 Regional geology ............................................................................................... 3 1.2.2 Lake Harbour Group marble and calc-silicate rocks.......................................... 4 1.3 Broad research questions and significance ........................................................................ 5 1.3.1 Part 1 - Corundum .............................................................................................. 6 1.3.2 Part 2 - Spinel ..................................................................................................... 7 1.3.3 Part 3 - Lapis lazuli ............................................................................................ 8 1.4 Background on gem corundum .......................................................................................... 9 1.4.1 Geology of gem corundum deposits .................................................................. 9 1.4.2 Colour of corundum ......................................................................................... 10 1.4.3 Previous studies on Kimmirut sapphire deposits ............................................. 10 1.5 Background on gem spinel .............................................................................................. 11 1.5.1 Geology of gem spinel deposits ....................................................................... 11 1.5.2 Colour of spinel ................................................................................................ 12 1.5.3 Previous studies on Lake Harbour Group spinel.............................................. 13 1.6 Background on lapis lazuli .............................................................................................. 14 1.6.1 Lazurite vs. haüyne .......................................................................................... 14 1.6.2 Geology of lapis lazuli deposits ....................................................................... 14 1.6.3 Geology of the Soper River lapis lazuli occurrence ........................................ 16 1.7 Figures ............................................................................................................................. 17 ix  Chapter 2. Methods ....................................................................................................................... 19 2.1 Sapphire ........................................................................................................................... 19 2.1.1 Petrography and Mineral Identification ........................................................... 19 2.1.2 Chemical analysis............................................................................................. 19 2.1.3 Boron isotopes .................................................................................................. 20 2.1.4 Oxygen isotopes of corundum ......................................................................... 21 2.1.5 Whole-rock geochemistry ................................................................................ 21 2.1.6 Radiometric dating ........................................................................................... 22 2.2 Spinel ............................................................................................................................... 23 2.2.1 Sampling, petrography and mineral identification ........................................... 23 2.2.2 Electron probe microanalysis ........................................................................... 24 2.2.3 Whole-rock geochemistry ................................................................................ 24 2.2.4 Thermodynamic modelling .............................................................................. 25 2.3 Lapis lazuli ....................................................................................................................... 26 2.3.1 Whole-rock geochemistry ................................................................................ 26 Chapter 3. Origin of scapolite-hosted sapphire (corundum) near Kimmirut ................................ 27 3.1 Results summary .............................................................................................................. 27 3.2 Chapter introduction ........................................................................................................ 28 3.3 Exploration History.......................................................................................................... 29 3.4 Results .............................................................................................................................. 30 3.4.1 Outcrop descriptions ........................................................................................ 30 3.4.2 Petrography and mineral compositions ............................................................ 30 3.4.3 Whole rock composition – Beluga deposit ...................................................... 34 3.4.4 Boron and oxygen isotope compositions ......................................................... 35 3.4.5 Zircon U-Pb geochronology ............................................................................. 35 3.4.6 Ar-Ar ages of mica and estimate of Tc ............................................................. 36 3.5 Discussion ........................................................................................................................ 36 3.5.1 Paragenetic sequence and metamorphic history .............................................. 36 3.5.2 Controls on corundum genesis ......................................................................... 41 3.5.3 A possible magmatic origin? ............................................................................ 42 3.5.4 Nature of the protolith ...................................................................................... 43 3.5.5 On evaporites ................................................................................................... 46 3.5.6 Implications for gem corundum exploration .................................................... 47 3.6 Conclusions ...................................................................................................................... 48 3.7 Tables ............................................................................................................................... 51 3.8 Figures ............................................................................................................................. 61 x  Chapter 4. Metacarbonate-hosted spinel on Baffin Island: Insights into the origin of gem spinel and cobalt-blue spinel ................................................................................................................... 81 4.1 Results summary .............................................................................................................. 81 4.2 Chapter introduction ........................................................................................................ 82 4.3 Results .............................................................................................................................. 83 4.3.1 Petrography ...................................................................................................... 83 4.3.2 Mineral compositions ....................................................................................... 97 4.3.3 Whole-rock compositions .............................................................................. 103 4.3.4 Pseudo-sections .............................................................................................. 109 4.4 Calculations ................................................................................................................... 111 4.4.1 Sedimentary protolith composition estimation: Method and underlying assumptions .................................................................................................. 111 4.4.2 Estimate of “expected” minor metal content ................................................. 112 4.5 Discussion ...................................................................................................................... 113 4.5.1 Origin of spinel occurrences, parageneses, P-T conditions and timing ......... 113 4.5.2 Metasedimentary spinel protoliths ................................................................. 119 4.5.3 Geochemical factors in spinel genesis ........................................................... 123 4.5.4 Origin of cobalt enrichment at cobalt-blue spinel occurrences ...................... 125 4.5.5 Controls on spinel color ................................................................................. 133 4.5.6 Exploration criteria......................................................................................... 135 4.6 Conclusions .................................................................................................................... 136 4.6.1 Origin of spinel occurrences .......................................................................... 136 4.6.2 Metasedimentary protoliths ............................................................................ 136 4.6.3 Geochemical factors in spinel genesis ........................................................... 137 4.6.4 Origin of cobalt enrichment at Qila and Trailside ......................................... 137 4.6.5 Controls on spinel color ................................................................................. 138 4.6.6 Exploration criteria......................................................................................... 138 4.7 Tables ............................................................................................................................. 139 4.8 Figures ........................................................................................................................... 175 Chapter 5. Soper River lapis lazuli: Protolith and effect on quality ........................................... 203 5.1 Chapter introduction ...................................................................................................... 203 5.2 Results ............................................................................................................................ 203 5.2.1 Mineralogical composition............................................................................. 203 5.2.2 Whole rock major element composition ........................................................ 203 5.2.3 Whole rock trace element composition .......................................................... 204 5.3 Discussion ...................................................................................................................... 204 5.3.1 Evaporitic origin and protolith ....................................................................... 204 5.3.2 Composition of protolith vs. gem quality of lapis lazuli ................................ 205 xi  5.4 Conclusions .................................................................................................................... 206 5.5 Figures ........................................................................................................................... 207 Chapter 6. General deposit model and exploration criteria ........................................................ 208 6.1 Fundamental assumptions about LHG metacarbonate protoliths .................................. 208 6.2 Gem deposit model ........................................................................................................ 209 6.2.1 Sedimentary protoliths and metasomatic occurrences ................................... 209 6.2.2 Geochemical controls on gem mineral genesis .............................................. 210 6.2.3 Metamorphic controls on gem mineral genesis ............................................. 211 6.2.4 Controls on chromophore concentrations ...................................................... 213 6.3 Exploration criteria & methods...................................................................................... 215 6.3.1 Kimmirut-type sapphire deposits ................................................................... 215 6.3.2 Gem spinel deposits ....................................................................................... 216 6.3.3 Lapis lazuli deposits ....................................................................................... 217 6.3.4 Exploration using aerial surveys .................................................................... 217 6.4 Conclusions .................................................................................................................... 218 6.5 Figures ........................................................................................................................... 220 Chapter 7. Conclusions and future work ..................................................................................... 221 References ................................................................................................................................... 225 Appendix A ................................................................................................................................. 243    xii  List of Tables TABLE 3.1: AVERAGE COMPOSITION OF PHLOGOPITE AND MUSCOVITE FROM THE BELUGA AND BOWHEAD CALC-SILICATE PODS. AFTER DZIKOWSKI (2013). .. 51 TABLE 3.2: AVERAGE COMPOSITION OF PLAGIOCLASE AND NEPHELINE FROM THE BELUGA AND BOWHEAD CALC-SILICATE PODS. AFTER DZIKOWSKI (2013). ................................................................................................................................... 52 TABLE 3.3: AVERAGE COMPOSITION OF DIOPSIDE FROM THE BELUGA AND BOWHEAD CALC-SILICATE PODS. AFTER DZIKOWSKI (2013). ............................. 53 TABLE 3.4: COMPOSITION OF ZIRCON FROM THE PHLOGOPITE-OLIGOCLASE PORTION OF BELUGA CALC-SILICATE ROCK. .......................................................... 53 TABLE 3.5: AVERAGE COMPOSITION OF SCAPOLITE FROM BELUGA AND BOWHEAD CALC-SILICATE PODS WITH STANDARD DEVIATION AND MINIMUM/MAXIMUM ME% COMPOSITIONS. SEE DZIKOWSKI (2013) FOR FULL DATA SET. .......................................................................................................................... 54 TABLE 3.6: AVERAGE COMPOSITION OF OXY-DRAVITE AT THE BELUGA OCCURRENCE. ................................................................................................................... 55 TABLE 3.7: WHOLE ROCK MAJOR ELEMENT COMPOSITION, BELUGA OCCURRENCE. AFTER DZIKOWSKI (2013). ................................................................. 56 TABLE 3.8: WHOLE ROCK TRACE ELEMENT COMPOSITION, BELUGA OCCURRENCE. AFTER DZIKOWSKI (2013). ................................................................. 57 TABLE 3.9: BORON ISOTOPE COMPOSITION OF OXY-DRAVITE FROM THE BELUGA OCCURRENCE. ................................................................................................................... 58 TABLE 3.10: U-PB ANALYTICAL DATA FOR ZIRCON FROM THE BELUGA OCCURRENCE. ................................................................................................................... 59 TABLE 3.11: ESTIMATED COOLING RATE AND CLOSURE TEMPERATURES FOR PHLOGOPITE AND MUSCOVITE AT THE BELUGA OCCURRENCE USING DIFFERENT DETERMINATIONS OF ACTIVATION ENERGY (EA) AND FREQUENCY FACTOR (D0) FOR PHLOGOPITE (PHL) AND MUSCOVITE (MS) AND CALCULATED IN CLOSURE V1.2................................................................................... 60 TABLE 4.1: SPINEL OCCURRENCES STUDIED .................................................................. 139 TABLE 4.2A: SOUTH TO NORTH SECTION OF LITHOLOGIES SURROUNDING THE WEST EXTREMITY OF THE DIOPSIDITE BAND, SPINEL ISLAND, MARKHAM BAY. ................................................................................................................................... 140 xiii  TABLE 4.2B: SOUTH TO NORTH SECTION OF LITHOLOGIES SURROUNDING THE SPINEL-BEARING ZONE OF THE DIOPSIDITE BAND, SPINEL ISLAND, MARKHAM BAY. ............................................................................................................. 140 TABLE 4.3A: AVERAGE COMPOSITION OF SPINEL FROM MARKHAM BAY, WADDELL BAY, AND HALL PENINSULA. NORMALIZED TO 32 OXYGEN ATOMS PER FORMULA UNIT. ..................................................................................................... 141 TABLE 4.3B: AVERAGE COMPOSITION OF SPINEL FROM GLENCOE ISLAND. NORMALIZED TO 32 OXYGEN ATOMS PER FORMULA UNIT. ............................. 142 TABLE 4.3C: AVERAGE COMPOSITION OF SPINEL FROM SOPER LAKE AND SOPER RIVER AREA, NEAR KIMMIRUT. NORMALIZED TO 32 OXYGEN ATOMS PER FORMULA UNIT. ............................................................................................................. 143 TABLE 4.3D: AVERAGE COMPOSITION OF SPINEL FROM QILA AND TRAILSIDE, KIMMIRUT AREA. NORMALIZED TO 32 OXYGEN ATOMS PER FORMULA UNIT.............................................................................................................................................. 144 TABLE 4.4: AVERAGE COMPOSITION OF FORSTERITE ASSOCIATED WITH SPINEL, SOUTHERN BAFFIN ISLAND. NORMALIZED TO 4 OXYGEN ATOMS PER FORMULA UNIT. ............................................................................................................. 145 TABLE 4.5A: COMPOSITION OF DIOPSIDE ASSOCIATED WITH SPINEL, MARKHAM BAY AND GLENCOE ISLAND. NORMALIZED TO 6 OXYGEN ATOMS PER FORMULA UNIT. ............................................................................................................. 146 TABLE 4.5B: AVERAGE COMPOSITION OF DIOPSIDE ASSOCIATED WITH SPINEL, KIMMIRUT AREA, WADDELL BAY, AND HALL PENINSULA. NORMALIZED TO 6 OXYGEN ATOMS PER FORMULA UNIT. .................................................................... 147 TABLE 4.6A: AVERAGE COMPOSITION OF AMPHIBOLE FROM MARKHAM BAY, GLENCOE ISLAND, AND HALL PENINSULA. RESULTS CALCULATED USING THE SPREADSHEET OF LOCOCK (2014) FOR 24 ANIONS. ...................................... 148 TABLE 4.6B: AVERAGE COMPOSITION OF AMPHIBOLE FROM THE KIMMIRUT AREA. RESULTS CALCULATED USING THE SPREADSHEET OF LOCOCK (2014) FOR 24 ANIONS. ............................................................................................................... 149 TABLE 4.6C: AVERAGE COMPOSITION OF AMPHIBOLE FROM THE KIMMIRUT AREA. RESULTS CALCULATED USING THE SPREADSHEET OF LOCOCK (2014) FOR 24 ANIONS. ............................................................................................................... 150 TABLE 4.7A: AVERAGE COMPOSITION OF PHLOGOPITE FROM MARKHAM BAY, WADDELL BAY, AND HALL PENINSULA. NORMALIZED ON THE BASIS OF 12 ANIONS PER FORMULA UNIT. ..................................................................................... 151 xiv  TABLE 4.7B: AVERAGE COMPOSITION OF PHLOGOPITE FROM GLENCOE ISLAND. NORMALIZED ON THE BASIS OF 12 ANIONS PER FORMULA UNIT. .................. 152 TABLE 4.7C: AVERAGE COMPOSITION OF PHLOGOPITE FROM THE SOPER LAKE MINE, KIMMIRUT AREA. NORMALIZED ON THE BASIS OF 12 ANIONS PER FORMULA UNIT. ............................................................................................................. 153 TABLE 4.7D: AVERAGE COMPOSITION OF PHLOGOPITE FROM QILA, KIMMIRUT AREA. NORMALIZED ON THE BASIS OF 12 ANIONS PER FORMULA UNIT. ...... 154 TABLE 4.7E: AVERAGE COMPOSITION OF PHLOGOPITE FROM TRAILSIDE, KIMMIRUT AREA. NORMALIZED ON THE BASIS OF 12 ANIONS PER FORMULA UNIT. .................................................................................................................................. 155 TABLE 4.8: AVERAGE COMPOSITION OF HUMITE FROM SOPER FALLS. NORMALIZED ON THE BASIS OF 3 SI ATOMS PER FORMULA UNIT. ................. 156 TABLE 4.9: AVERAGE COMPOSITION OF SCAPOLITE FROM SPINEL-BEARING CALC-SILICATE ROCK AT QILA. NORMALIZED TO AL+SI=12. ............................ 157 TABLE 4.10: AVERAGE COMPOSITION OF MUSCOVITE FROM SPINEL-BEARING SILICATE ROCK AT TRAILSIDE. NORMALIZED TO 12 ANIONS PER FORMULA UNIT. .................................................................................................................................. 158 TABLE 4.11A: AVERAGE COMPOSITION OF CALCITE ASSOCIATED WITH SPINEL FROM MARKHAM BAY AND GLENCOE ISLAND. ................................................... 159 TABLE 4.11B: AVERAGE COMPOSITION OF CALCITE ASSOCIATED WITH SPINEL FROM KIMMIRUT, WADDELL BAY, AND THE HALL PENINSULA. ..................... 159 TABLE 4.12A: AVERAGE COMPOSITION OF DOLOMITE AND DOLOMITE EXSOLUTION IN CALCITE ASSOCIATED WITH SPINEL FROM BAFFIN ISLAND.............................................................................................................................................. 160 TABLE 4.12B: AVERAGE COMPOSITION OF DOLOMITE AND DOLOMITE EXSOLUTION IN CALCITE ASSOCIATED WITH SPINEL FROM BAFFIN ISLAND.............................................................................................................................................. 160 TABLE 4.13: AVERAGE COMPOSITION OF PYRRHOTITE AND PYRITE ASSOCIATED WITH SPINEL ON BAFFIN ISLAND. ............................................................................. 161 TABLE 4.14A: MAJOR ELEMENT COMPOSITION OF WHOLE ROCK SAMPLES FROM SPINEL ISLAND, GLENCOE ISLAND, AND THE HALL PENINSULA. .................... 162 TABLE 4.14B: MAJOR ELEMENT COMPOSITION OF WHOLE ROCK SAMPLES FROM SOPER RIVER, SOPER LAKE, AND SOPER FALLS, KIMMIRUT AREA. ................ 163 xv  TABLE 4.14C: MAJOR ELEMENT COMPOSITION OF WHOLE ROCK SAMPLES FROM QILA AND TRAILSIDE, KIMMIRUT AREA. ................................................................ 164 TABLE 4.15A: TRACE ELEMENT CONCENTRATIONS (µG/G) OF WHOLE ROCK SAMPLES FROM SPINEL ISLAND, GLENCOE ISLAND, AND THE HALL PENINSULA. BELOW DETECTION LIMIT: GE (< 5 µG/G). ....................................... 165 TABLE 4.15B: TRACE ELEMENT CONCENTRATIONS (µG/G) OF WHOLE ROCK SAMPLES FROM SOPER RIVER, SOPER LAKE, AND SOPER FALLS, KIMMIRUT AREA. BELOW DETECTION LIMIT (µG/G): GE (5), CD (0.5). ................................... 166 TABLE 4.15C: TRACE ELEMENT CONCENTRATIONS (µG/G) OF WHOLE ROCK SAMPLES FROM QILA, KIMMIRUT AREA. BELOW DETECTION LIMIT (µG/G): GE (5); AG AND CD (0.5); RE (0.001). .................................................................................. 167 TABLE 4.15D: TRACE ELEMENT CONCENTRATIONS (µG/G) OF WHOLE ROCK SAMPLES FROM TRAILSIDE, KIMMIRUT AREA. BELOW DETECTION LIMIT (µG/G): GE (5); AG AND CD (0.5); RE (0.001). .............................................................. 168 TABLE 4.16A: ESTIMATED PROTOLITH COMPOSITION OF METACARBONATE SAMPLES FROM MARKHAM BAY, GLENCOE ISLAND. ......................................... 169 TABLE 4.16B: ESTIMATED PROTOLITH COMPOSITION OF METACARBONATE SAMPLES FROM HALL PENINSULA AND PART OF THE KIMMIRUT AREA. ..... 170 TABLE 4.16C: ESTIMATED PROTOLITH COMPOSITION OF METACARBONATE SAMPLES FROM SOPER RIVER AND QILA, KIMMIRUT AREA. ............................ 171 TABLE 4.16D: ESTIMATED PROTOLITH COMPOSITION OF METACARBONATE SAMPLES FROM TRAILSIDE, KIMMIRUT AREA. ..................................................... 172 TABLE 4.17A: “EXPECTED” TRACE ELEMENT COMPOSITION OF METASEDIMENT SAMPLES CALCULATED BASED ON MIXING OF SEDIMENTARY ROCK AVERAGES AND COMPARISON TO ACTUAL VALUES METASEDIMENTS (CALCULATED USING AVERAGES FOR SHALE/CLAY, SANDSTONE, AND LIMESTONE AND CORRECTED FOR MASS LOSS DUE TO DEVOLATILIZATION; SEE TEXT). SAMPLES FROM MARKHAM BAY, GLENCOE ISLAND, HALL PENINSULA AND PART OF THE KIMMIRUT AREA. ................................................ 173 TABLE 4.17B: “EXPECTED” TRACE ELEMENT COMPOSITION OF METASEDIMENT SAMPLES CALCULATED BASED ON MIXING OF SEDIMENTARY ROCK AVERAGES AND COMPARISON TO ACTUAL VALUES METASEDIMENTS (CALCULATED USING AVERAGES FOR SHALE/CLAY, SANDSTONE, AND LIMESTONE AND CORRECTED FOR MASS LOSS DUE TO DEVOLATILIZATION; SEE TEXT). SAMPLES FROM QILA AND TRAILSIDE, KIMMIRUT AREA. ........... 174  xvi  List of Figures FIGURE 1.1: GEOLOGY OF SOUTHERN BAFFIN ISLAND AND LOCATION OF THE STUDY AREAS. CRUSTAL SUTURES SEPARATING STRUCTURAL DOMAINS ARE REPRESENTED BY DASHED LINES. BS: BERGERON SUTURE; SRS: SOPER RIVER SUTURE. MODIFIED AFTER ST-ONGE ET AL. (2000) AND BUTLER (2007); AGES FROM ST-ONGE ET AL. (2001). ............................................................................. 17 FIGURE 1.2: CLASSIFICATION OF GEM CORUNDUM DEPOSITS (AFTER SIMONET ET AL. 2008)............................................................................................................................... 18 FIGURE 1.3: P-T CONDITIONS FOR THE FORMATION OF CORUNDUM IN VARIOUS METAMORPHIC DEPOSITS (AFTER GIULIANI ET AL. 2014). .................................... 18 FIGURE 3.1: BEDROCK GEOLOGY MAP OF THE TRUE NORTH GEMS PROPERTY WITH MARKERS FOR SCAPOLITE, SPINEL, AND CORUNDUM OCCURRENCES FOUND DURING THE 2006 MAPPING SEASON. THE MOST IMPORTANT MINERALIZED AREAS ARE NAMED. SMALL PLUTONS CONSIST OF GRANITE OR ULTRAMAFIC PLUGS. UTM ZONE 19 V (NAD83). MAP COURTESY OF TRUE NORTH GEMS INC. ............................................................................................................ 61 FIGURE 3.2: (A: TOP) CORUNDUM (SAPPHIRE) GEMSTONES FROM THE KIMMIRUT OCCURRENCES. LEFT: COLORLESS SAPPHIRE, AQPIK OCCURRENCE, 2.50 AND 2.59 CT. TOP CENTRE: DEEP BLUE, EXTRA FINE SAPPHIRE, 1.17 CT, FROM THE BELUGA OCCURRENCE, AND A HEAT-TREATED, RICH BLUE 2.43 CT SAPPHIRE FROM AQPIK. RIGHT: YELLOW SAPPHIRE, BELUGA SOUTH OCCURRENCE, 1.09 CT AND 1.47 CT. PHOTOGRAPH COURTESY OF TRUE NORTH GEMS INC. (B: LOWER LEFT) LIGHTLY INCLUDED, LIGHT BLUE SAPPHIRE (7.81 CT) FROM THE AQPIK OCCURRENCE. PHOTOGRAPH BY BRAD WILSON. (C: LOWER RIGHT) DARK BLUE CORUNDUM CRYSTAL, 36 × 4 MM, IN CALC-SILICATE ROCK, AND A 1.17 CT SAPPHIRE GEMSTONE FROM THE BELUGA OCCURRENCE. PHOTOGRAPH BY BRAD WILSON, COURTESY OF TRUE NORTH GEMS INC............................................................................................................................ 62 FIGURE 3.3: CONTACT BETWEEN MARBLE AND SAPPHIRE-BEARING CALC-SILICATE ROCK AT THE BELUGA OCCURRENCE. NOTE THE VARIATION IN THE DISTRIBUTION OF LIGHT-COLORED AND DARK MINERAL ASSEMBLAGES WITHIN THE CALC-SILICATE POD, AND THE UNDULATE NATURE OF THE CONTACT WITH COARSELY CRYSTALLINE MARBLE. THE CALC-SILICATE POD IS SURROUNDED BY VERY COARSE-GRAINED MARBLE, AND NO FOLIATION IS APPARENT IN THE OUTCROP. ................................................................................... 63 FIGURE 3.4: CORUNDUM-BEARING CALC-SILICATE ROCK IN SITU AT THE BELUGA PIT. PHLOGOPITE-OLIGOCLASE (PHL + PL) INTERGROWTHS OCCUR NEAR PARTLY ALBITIZED SCAPOLITE (AB + SCP) AND ALBITE, MUSCOVITE, AND BLUE CORUNDUM (AB + MS + CRN). THE CORUNDUM- AND PHLOGOPITE-xvii  BEARING ASSEMBLAGES CONTAIN CALCITE. MINOR GRAPHITE (CG) IS PRESENT. ............................................................................................................................ 64 FIGURE 3.5: DIOPSIDE (DI) SURROUNDED BY PHLOGOPITE-OLIGOCLASE SYMPLECTITE AND COARSER ORIENTED INTERGROWTHS OF PHLOGOPITE (PHL), OLIGOCLASE (PL), AND CALCITE (CAL). BELUGA OCCURRENCE. CROSS-POLARIZED LIGHT. .......................................................................................................... 65 FIGURE 3.6: CONTACT BETWEEN COARSE-GRAINED SCAPOLITE (SCP), PHLOGOPITE-OLIGOCLASE-CALCITE (PHL-PL-CAL), AND THE CORUNDUM-BEARING ZONE. THE LATTER ZONE CONTAINS IDIOMORPHIC CORUNDUM (CRN) WITH ALBITE (AB), CALCITE (CAL), AND MUSCOVITE (MS) OF VARIABLE GRAIN SIZE. DARK ZONES OF FINE-GRAINED ALTERATION (ALT) CONSIST OF MIXTURES OF THE FOLLOWING MINERALS IN VARIABLE ABUNDANCE: ALBITE, CALCITE, MUSCOVITE, ANALCIME, PREHNITE, AND THOMSONITE. BELUGA OCCURRENCE. PLANE POLARIZED LIGHT. .................. 66 FIGURE 3.7: BELUGA CALC-SILICATE ROCK UNDER SHORTWAVE ULTRAVIOLET LIGHT SHOWING THE FLUORESCENT SCAPOLITE (YELLOW; SCP), VARIABLE DIOPSIDE-PHLOGOPITE-OLIGOCLASE ASSEMBLAGES (DI-PHL-PL), ALBITIZED SCAPOLITE (AB), AND ALBITE-CORUNDUM-MUSCOVITE ASSEMBLAGES (AB-CRN-MS). MINOR CALCITE IS PRESENT. PURPLE COLORATION IS AN ARTEFACT OF THE UV LIGHT SOURCE. ..................................................................... 67 FIGURE 3.8: THIN SCAPOLITE RIM BETWEEN A GRAIN OF NEPHELINE AND CALCITE, WHICH SUGGESTS THAT SCAPOLITE FORMED FROM THE REACTION OF CALCITE AND NEPHELINE AND POST-DATES THE NEPHELINE-BEARING MINERAL ASSEMBLAGE. BOWHEAD OCCURRENCE. CROSS-POLARIZED LIGHT................................................................................................................................... 68 FIGURE 3.9: U-PB CONCORDIA DIAGRAM FOR ZIRCON RECOVERED FROM THE PHLOGOPITE-RICH ASSEMBLAGE AT THE BELUGA OCCURRENCE CALCULATED USING MODEL 1 OF LUDWIG (2003). ERROR ELLIPSES REPRESENT 2Σ. SEE TABLE 10 AND TEXT FOR U-PB DATA AND SAMPLE DESCRIPTIONS. ................................................................................................................. 69 FIGURE 3.10 A-B: 40AR/39AR AGE SPECTRA OF PHLOGOPITE FROM THE BELUGA OCCURRENCE. BOX HEIGHTS ARE 2Σ. PLATEAU STEPS ARE FILLED AND REJECTED STEPS ARE OPEN. AFTER DZIKOWSKI (2013). ....................................... 70 FIGURE 3.10 C-D: 40AR/39AR AGE SPECTRA OF MUSCOVITE (MS; C, D) FROM THE BELUGA OCCURRENCE. BOX HEIGHTS ARE 2Σ. PLATEAU STEPS ARE FILLED AND REJECTED STEPS ARE OPEN. AFTER DZIKOWSKI (2013). ............................. 71 FIGURE 3.11A: MINERAL REACTIONS MODELED IN TWQ V1 (BERMAN 1988, 1991). BASED ON THE PARAGENETIC SEQUENCE OF BOWHEAD AND BELUGA CALC-SILICATE ROCK (PRESENT STUDY) AND METASEDIMENTS ON ALIGUQ ISLAND (BUTLER 2007), THE DIOPSIDE (DI), NEPHELINE (NE), AND K-xviii  FELDSPAR (KFS) ASSEMBLAGE IS SUGGESTED TO BE POSSIBLE AT M1A METAMORPHIC CONDITIONS BUT LOW XCO2 (SEE FIG. 3.11B). ALTERNATIVELY, THE FORMATION OF NEPHELINE AND DIOPSIDE FROM ALBITE (AB) AND DOLOMITE (DOL) IS CONSISTENT WITH REGIONAL P-T CONDITIONS BUT DIFFERS SIGNIFICANTLY FROM THE PARAGENETIC SEQUENCE AT BELUGA AND BOWHEAD. THE PEAK ASSEMBLAGE IS PARTLY REPLACED BY PHLOGOPITE (PHL), CALCITE (CAL), AND ALBITE AT CONDITIONS SLIGHTLY BELOW THAT OF M2. THE MEIONITE (ME) – NEPHELINE BREAK-DOWN REACTIONS ARE PROBABLY OVERESTIMATES SINCE THE MEASURED SCAPOLITE COMPOSITIONS CONTAIN SIGNIFICANT NA. THE LATTER MINERALS BREAK DOWN TO FORM CORUNDUM (CRN), CALCITE, AND ALBITE AT HIGHER TEMPERATURE THAN A SIMILAR REACTION FORMING AL-SILICATE, CALCITE, AND ALBITE. REGIONAL P-T CONDITIONS AFTER ST-ONGE ET AL. (2007). SEE TEXT FOR A DETAILED DISCUSSION OF MINERAL REACTIONS AND PARAGENETIC SEQUENCE IN THE CONTEXT OF REGIONAL METAMORPHISM. .............................................................. 72 FIGURE 3.11B: MINERAL REACTIONS MODELED IN TWQ (BERMAN 1988, 1991) MODIFIED AFTER BUTLER (2007) AT PEAK METAMORPHIC PRESSURE (8 KBAR, ST-ONGE ET AL. 2007). THE NEPHELINE-BEARING ASSEMBLAGE OCCURS AT LOW XCO2 AT THE PEAK METAMORPHIC TEMPERATURE (810°C, ST-ONGE ET AL. 2007). ...................................................................................................... 73 FIGURE 3.12: COMPARISON OF BELUGA CALC-SILICATE ROCK MAJOR ELEMENT COMPOSITION WITH SELECTED REPRESENTATIVE LAKE HARBOUR GROUP METASEDIMENTS (THÉRIAULT ET AL. 2001, BUTLER 2007), MONZOGRANITES (BUTLER 2007), LAZURITE/HAÜYNE-BEARING METAEVAPORITES (HOGARTH & GRIFFIN 1978), AVERAGES FOR PLATFORM SEDIMENTS (CARMICHAEL 1989), AND AVERAGES FOR SYENITE (NOCKOLDS 1954). ...................................... 74 FIGURE 3.13: ABUNDANCE OF SELECTED TRACE METALS IN BELUGA CALC-SILICATE ROCK COMPARED TO REPRESENTATIVE LHG METASEDIMENTS (THÉRIAULT ET AL. 2001), SEDIMENTS (CARMICHAEL 1989, MOINE ET AL. 1981 AND REFERENCES THEREIN), AND AVERAGE GRANITE (CARMICHAEL 1989). 75 FIGURE 3.14: ABUNDANCE OF V, CR, AND NI RELATIVE TO AL2O3. LINE REPRESENTS EXPECTED CONCENTRATION OF METALS RELATIVE TO ALUMINA BY VARYING AMOUNT OF ALLOCHTHONOUS DETRITUS BASED ON METAL-AL2O3 RATIOS FROM A SHALE AVERAGE (CARMICHAEL 1989). SELECTED EXAMPLES OF METAPELITE AND PSAMMITE FROM THE LAKE HARBOUR GROUP (THÉRIAULT ET AL. 2001) ARE INCLUDED FOR COMPARISON. ................................................................................................................... 76 FIGURE 3.15: CHONDRITE-NORMALIZED (VALUES OF TAYLOR & MCLENNAN 1985) REE TRACE ELEMENT PROFILE OF BELUGA CALC-SILICATE ROCK COMPARED TO ROCKS FROM THE LAKE HARBOUR GROUP (*BUTLER 2007, **THÉRIAULT ET AL. 2001). AFTER DZIKOWSKI (2013). .......................................... 77 xix  FIGURE 3.16: OXYGEN ISOTOPE COMPOSITION OF BELUGA CORUNDUM COMPARED TO VALUES FROM DIFFERENT TYPES OF GEM CORUNDUM DEPOSITS (MODIFIED AFTER GIULIANI ET AL. 2014). COLOR IN DIAMONDS REPRESENT THE COLOUR OF GEM-QUALITY CORUNDUM FROM DIFFERENT DEPOSITS (RED = RUBY, OTHERS = COLORED SAPPHIRE). DIAGRAM COURTESY OF G. GIULIANI. .......................................................................................... 78 FIGURE 3.17: AL-MG-CA DIAGRAM SHOWING THE DISTRIBUTION OF THE CALC-SILICATE ROCK FROM THE BELUGA CORUNDUM DEPOSIT RELATIVE TO DOMAINS FOR PLATFORM MARLS AND CLAY-SHALE, EVAPORITES, AND META-EVAPORITES (MODIFIED AFTER MOINE ET AL. 1981) AND IN COMPARISON TO MARBLES (OPEN SYMBOLS) AND INTERCALATED SCHISTS AND GNEISSES (FULL SYMBOLS) FROM RUBY-BEARING MARBLES IN CENTRAL AND SOUTH-EAST ASIA (AFTER GARNIER ET AL. 2008). DIAGRAM COURTESY OF G. GIULIANI. .......................................................................................... 79 FIGURE 18: BORON ISOTOPE COMPOSITION OF BELUGA OXY-DRAVITE COMPARED TO TOURMALINE FROM DIFFERENT ENVIRONMENTS. SOURCES OF DATA: (1) SWIHART & MOORE 1989; (2) PALMER 1991; (3) SERENDIBITE-RICH CALC-SILICATE ROCK INTERPRETED TO BE METAMORPHOSED ILLITE LAYER DEPOSITED IN A HYPERSALINE ENVIRONMENT (GREW ET AL. 1991); (4) SERENDIBITE-RICH CALC-SILICATE ROCK SIMILAR TO THE LATTER, BUT WITH BOTH PROGRADE AND RETROGRADE TOURMALINE (BELLEY ET AL. 2014). .................................................................................................................................... 80 FIGURE 4.1: LOCATION OF THE MARKHAM BAY SPINEL LOCALITIES (1A AND 1B). GEOLOGY AFTER BLACKADAR (1967) AND BUTLER (2007). ONLY THE SOUTHEASTERN PART OF “SPINEL ISLAND” WAS INSPECTED IN DETAIL IN THE PRESENT STUDY. DOTTED LINES REPRESENT THE EXTENT OF EXPOSURE AT LOW TIDE. .................................................................................................................. 175 FIGURE 4.2: SPINEL-BEARING DIOPSIDITE BAND (CA. 2.5 M THICK) SEEN LOOKING WEST ON SPINEL ISLAND, NORTH OF MACDONALD ISLAND, MARKHAM BAY.............................................................................................................................................. 175 FIGURE 4.3: (A: TOP LEFT) MAIN AREA OF THE DIOPSIDITE BAND AT SPINEL ISLAND. NOTE THE UNEVEN DISTRIBUTION OF PHLOGOPITE (PHL)-RICH PORTIONS. SPINEL (SPL) CRYSTALS OCCUR IN THE LOWER PART OF THE OUTCROP, WHERE THE DIOPSIDITE IS POORER IN PHLOGOPITE. PALE-COLORED CALCITE (CAL), DIOPSIDE (DI), AND PHLOGOPITE OCCUR ON THE HANGING WALL OF THE DIOPSIDITE, WHERE IT PROBABLY GRADES INTO MARBLE. (B: TOP RIGHT) DARK BLUE, EUHEDRAL SPINEL CRYSTALS UP TO 4 CM IN SIZE IN WHITE CALCITE, GREY DIOPSIDE, AND BROWN PHLOGOPITE. (C: BOTTOM) SPINEL (SPL) PARTLY REPLACED BY CORUNDUM (CRN) AND CLINOCHLORE (CHL). TRACE GAHNITE (GAH). LOWER PART OF IMAGE CONTAINS RETROGRADE MINERALS MARGARITE (MRG), ZOISITE OR xx  CLINOZOISITE (ZS/CZS), AND TREMOLITE (TR). SPINEL ISLAND, MARKHAM BAY. BACKSCATTERED ELECTRON IMAGE. ........................................................... 176 FIGURE 4.4: YELLOWISH-WHITE AMPHIBOLE-BEARING CALCITE VEINS AND DARK AMPHIBOLE VEINLETS CROSS-CUTTING DIOPSIDITE. UNNAMED ISLAND, MARKHAM BAY. ............................................................................................ 177 FIGURE 4.5: LOCATION OF SPINEL OCCURRENCES ON GLENCOE ISLAND: (2A) MAIN OCCURRENCE, (2B) MARBLE BEACH, (2C) CONTACT, AND (2D) MARBLE GULCH. ALL SPINEL OCCURRENCES ARE HOSTED IN LAKE HARBOUR GROUP (LHG) MARBLES AND CALC-SILICATE ROCKS. MODIFIED AFTER ST-ONGE ET AL. (1999B)......................................................................................................................... 177 FIGURE 4.6: (A) MAIN SPINEL OCCURRENCE AT GLENCOE ISLAND IN A 2-METER THICK BAND OF MARBLE. (B) CALC-SILICATE POD (DISJOINTED BOUDINS) AT THE GLENCOE MAIN OCCURRENCE CONTAINING SPINEL (SPL), DIOPSIDE (DI), CALCITE (CAL), AND PHLOGOPITE (PHL) HOSTED IN MARBLE MATRIX. ....... 178 FIGURE 4.7: EMBAYED GRAIN BOUNDARIES BETWEEN FORSTERITE (FO), PARGASITE (PRG), AND DIOPSIDE (DI) WITH APATITE (AP) AND RETROGRADE SERPENTINE (SERP). MARBLE GULCH OCCURRENCE, GLENCOE ISLAND. ..... 179 FIGURE 4.8: LOCATION OF SPINEL OCCURRENCES NEAR KIMMIRUT SAMPLED FOR THIS STUDY (REFER TO TABLE 4.1 FOR LOCALITIES). MODIFIED AFTER ST-ONGE ET AL. (1999A)....................................................................................................... 180 FIGURE 4.9: SPINEL IN DOLOMITIC MARBLE. SOPER RIVER SPINEL OCCURRENCE, KIMMIRUT AREA. ........................................................................................................... 181 FIGURE 4.10: SPINEL-BEARING HUMITITE PODS (HU) IN WEATHERED MARBLE OUTCROP. SOPER FALLS SPINEL OCCURRENCE, KIMMIRUT AREA. ................ 182 FIGURE 4.11: (A: TOP LEFT) SPINEL (SPL) AND WHITE CARBONATE IN HUMITE (HU). (B: TOP RIGHT) TEXTURAL RELATIONSHIP BETWEEN PARGASITE (PRG) AND HUMITE (HU) SUGGESTING REPLACEMENT OF THE LATTER BY THE FORMER. SAMPLE CONTAINS TALC AND MAGNESITE (MGS). BACKSCATTERED ELECTRON IMAGE. (C: BOTTOM) DOLOMITE (DOL) BORDERING MAGNESITE (MGS) AND HUMITE (HU). SPINEL (SPL) IS SURROUNDED BY A THIN RIM OF CLINOCHLORE (CHL). SOPER FALLS SPINEL OCCURRENCE, KIMMIRUT AREA. .............................................................................. 183 FIGURE 4.12: (A) PHLOGOPITE (PHL) PORPHYROBLAST IN COARSE-GRAINED FORSTERITE (FO) AND CARBONATE ROCK IN THE ROCK FACE AT THE SOPER LAKE MICA MINE, KIMMIRUT AREA. (B) PURPLE SPINEL (SPL) WITH PHLOGOPITE (PHL) AND FORSTERITE (FO) IN THE SOPER LAKE MICA PIT, KIMMIRUT AREA. ........................................................................................................... 184 xxi  FIGURE 4.13: (A: TOP) COBALT-BLUE SPINEL OUTCROP AT THE QILA OCCURRENCE: [3E-1] CALC-SILICATE POD; [3E-2] PARGASITE-CALCITE ROCK; [3E-3] MARBLE. (B: BOTTOM) VIVID BLUE SPINEL WITH WHITE CARBONATE IN CALC-SILICATE ROCK COMPOSED OF GREEN PARGASITE WITH SUBORDINATE GREYISH SCAPOLITE. QILA OCCURRENCE [ROCK UNIT 3E-1], KIMMIRUT AREA. ........................................................................................................... 185 FIGURE 4.14: FORSTERITE (FO) WITH A DIOPSIDE (DI) INTERIOR CORONA AND PARTIAL PARGASITE (PRG) EXTERIOR CORONA IN CALCITE (CAL) AND DOLOMITE (DOL) MARBLE. TRACE APATITE (AP). BACKSCATTERED ELECTRON IMAGE. QILA COBALT-SPINEL MARBLE (SAMPLE 3E-3-A), KIMMIRUT AREA. ........................................................................................................... 186 FIGURE 4.15: SILICATE-RICH SPINEL-BEARING ROCK (E.G., SAMPLE 3F-1), THE PREDOMINANT SPINEL-BEARING UNIT AT THE TRAILSIDE OCCURRENCE, KIMMIRUT AREA. IT IS COMPOSED OF FINE-GRAINED MUSCOVITE (MS) PSEUDOMORPHS AFTER AN UNKNOWN MINERAL, COARSE-GRAINED PHLOGOPITE (PHL), CALCITE (CAL), AND SPINEL (SPL). ..................................... 187 FIGURE 4.16: SPINEL (SPL) PARTLY ALTERED TO CORUNDUM (CRN) AND CLINOCHLORE (CHL). PHLOGOPITE (PHL) HAS RIMS OF CLINOCHLORE. WITH ACCESSORY PYRITE (PY) AND HOSTED IN CALCITE (CAL). BACKSCATTERED ELECTRON IMAGE. TRAILSIDE OCCURRENCE (ZONE 3F-1), KIMMIRUT AREA.............................................................................................................................................. 188 FIGURE 4.17: COBALT-BLUE SPINEL GEMSTONE, 0.16 CARATS (0.032 GRAMS) FROM THE TRAILSIDE OCCURRENCE, KIMMIRUT AREA. B.S. WILSON PHOTO. AFTER WILSON (2014). ................................................................................................................ 189 FIGURE 4.18: FOLIATED ROCK COMPOSED OF PALE YELLOW CALCITE, PALE BROWN PHLOGOPITE, AND COBALT-BLUE SPINEL (SAMPLE 3F-2) FOUND AS FLOAT ABOVE THE MAIN OUTCROP. TRAILSIDE OCCURRENCE, KIMMIRUT AREA. ................................................................................................................................. 189 FIGURE 4.19: APPARENT REPLACEMENT OF FORSTERITE (FO) BY DIOPSIDE (DI). BACKSCATTERED ELECTRON IMAGE. SAMPLE 3F-M1 FROM THE VICINITY OF THE TRAILSIDE OCCURRENCE, KIMMIRUT AREA. ................................................ 190 FIGURE 4.20: T-X(CO2) PSEUDO-SECTION FOR SAMPLE 2A-SPL-2, DIOPSIDE-CALCITE-PHLOGOPITE-SPINEL ROCK FROM THE GLENCOE MAIN OCCURRENCE, GLENCOE ISLAND. MINERAL ASSEMBLAGE “1” CONSISTS OF FORSTERITE (FO), DIOPSIDE (DI), SPINEL (SPL), PHLOGOPITE (PHL), CALCITE (CAL) AND DOLOMITE (DOL). GREY SHADING INDICATES CONDITIONS AT WHICH SPINEL IS STABLE. M1A REPRESENTS PEAK METAMORPHIC CONDITIONS AFTER ST-ONGE ET AL. (2007). ........................................................... 191 FIGURE 4.21: T-X(CO2) PSEUDO-SECTION FOR SAMPLE 3A-1, REPRESENTATIVE MARBLE FROM THE SOPER RIVER SPINEL OCCURRENCE NEAR KIMMIRUT xxii  CONTAINING CALCITE (CAL) AND DOLOMITE (DOL) WITH SUBORDINATE SPINEL (SPL), PARGASITE (PRG), PHLOGOPITE (PHL), AND FINE-GRAINED REPLACEMENTS OF DIOPSIDE (DI) AND DOLOMITE AFTER AN UNKNOWN MINERAL, PROBABLY FORSTERITE (FO). MINERAL ASSEMBLAGE “1” CONSISTS OF FO, PRG, SPL, PHL, ASP, CAL AND DOL. ASPIDOLITE (ASP; NA-END-MEMBER OF PHLOGOPITE) IS PREDICTED TO OCCUR, BUT THE SOLUTION MODEL FOR PHLOGOPITE/BIOTITE DOES NOT ACCOUNT FOR SUBSTITUTION OF NA FOR K. GREY SHADING INDICATES CONDITIONS AT WHICH SPINEL IS STABLE. M1A REPRESENTS PEAK METAMORPHIC CONDITIONS AFTER ST-ONGE ET AL. (2007). ........................................................... 192 FIGURE 4.22: T-X(CO2) PSEUDO-SECTION FOR SAMPLE 3E-3-A, SPINEL-BEARING MARBLE FROM THE QILA OCCURRENCE NEAR KIMMIRUT. MINERAL ASSEMBLAGE “1” CONSISTS OF FORSTERITE (FO), PARGASITE (PRG), DIOPSIDE (DI), SPINEL (SPL), PHLOGOPITE (PHL), CALCITE (CAL) AND DOLOMITE (DOL). GREY SHADING INDICATES CONDITIONS AT WHICH SPINEL IS STABLE. M1A REPRESENTS PEAK METAMORPHIC CONDITIONS AFTER ST-ONGE ET AL. (2007). ........................................................................................................ 193 FIGURE 4.23: T-X(CO2) PSEUDO-SECTION FOR SAMPLE 3F-2, PHLOGOPITE-CARBONATE-SPINEL ROCK FROM THE TRAILSIDE OCCURRENCE NEAR KIMMIRUT. MINERAL ASSEMBLAGE “1” CONSISTS OF SCAPOLITE (SCP), SPINEL (SPL), PHLOGOPITE (PHL), NEPHELINE (NE), CALCITE (CAL), AND DOLOMITE (DOL). ASSEMBLAGE “2” CONSISTS OF SCP, SPL, PHL, CAL, AND DOL. ALSO PLOTTED ARE GEIKIELITE (GK), PARGASITE (PRG), AND CORUNDUM (CRN) BEARING ASSEMBLAGES. ASPIDOLITE (ASP; NA-END-MEMBER OF PHLOGOPITE) IS PREDICTED TO OCCUR, BUT THE SOLUTION MODEL FOR PHLOGOPITE/BIOTITE DOES NOT ACCOUNT FOR SUBSTITUTION OF NA FOR K. GREY SHADING INDICATES CONDITIONS AT WHICH SPINEL IS STABLE. M1A REPRESENTS PEAK METAMORPHIC CONDITIONS AFTER ST-ONGE ET AL. (2007). ........................................................................................................ 194 FIGURE 4.24: ALUMINIUM AND SILICON CONTENT OF LHG METASEDIMENTARY ROCKS AND REFERENCE SEDIMENTARY ROCKS (KAOLINITE-RICH CLAYSTONE #AR-33, CLAYSTONE AND SILTSTONE AVERAGES AFTER LÓPEZ ET AL. 2005; AVERAGE SHALE/CLAY AFTER PARKER 1967; AVERAGE SANDSTONE AFTER TUREKIAN & WEDEPOHL 1961). DASHED LINES REPRESENT AL/SI FOR THE REFERENCE SEDIMENT VALUES USED IN PROTOLITH COMPOSITION ESTIMATE CALCULATIONS. .................................... 195 FIGURE 4.25: TERNARY PLOT OF THE ESTIMATED PROTOLITH SEDIMENTARY COMPOSITION (WT. %) FOR LHG SPINEL-BEARING ROCKS, CALC-SILICATE ROCKS, MARBLE, BELUGA SAPPHIRE CALC-SILICATE ROCK, AND SOPER RIVER LAPIS LAZULI. FOR LAPIS LAZULI, THE EVAPORITE COMPONENT IS INCLUDED WITH CARBONATE. .................................................................................. 196 xxiii  FIGURE 4.26: CAO AND MGO CONTENT OF LHG CALC-SILICATE ROCKS, MARBLES, BELUGA SAPPHIRE-BEARING CALC-SILICATE ROCK, AND SOPER RIVER LAPIS LAZULI. ............................................................................................................................. 197 FIGURE 4.27: (A) MG-AL-SI TERNARY DIAGRAM OF LHG METACARBONATE ROCKS AND PSAMMITE (RELATIVE MOL. %); (B) MG-AL-K TERNARY DIAGRAM OF LHG METACARBONATE ROCKS AND PSAMMITE (RELATIVE MOL. %). .......... 198 FIGURE 4.28: WHOLE ROCK CO/FE PLOTTED AGAINST NI/FE TO ILLUSTRATE THE RANGE OF QILA AND TRAILSIDE COMPARED TO OTHER LAKE HARBOUR GROUP METASEDIMENTS (MOSTLY CALC-SILICATE METACARBONATE AND MARBLE) AND AN EXAMPLE SET OF SILICICLASTIC SEDIMENTARY ROCKS (LÓPEZ ET AL. 2005). LHG SAMPLES WITH CO OR NI BELOW DETECTION LIMIT ARE EXCLUDED FROM THE DIAGRAM. .................................................................... 199 FIGURE 4.29: WHOLE ROCK CONCENTRATIONS OF COBALT AND IRON ILLUSTRATING THE HIGH LEVEL OF CO ENRICHMENT RELATIVE TO FE IN ROCKS FROM QILA AND TRAILSIDE IN COMPARISON TO OTHER LAKE HARBOUR GROUP METASEDIMENTS ANALYZED IN THE PRESENT STUDY. SAMPLES WITH CO BELOW DETECTION LIMIT ARE EXCLUDED FROM THE DIAGRAM. ........................................................................................................................ 200 FIGURE 4.30: WHOLE ROCK CO/AL VS. CO/FE IN LAKE HARBOUR GROUP METASEDIMENTS ILLUSTRATING THE WIDE RANGE IN CO/AL AT QILA AND TRAILSIDE DUE TO LOW AL CONTENTS IN QILA CO-RICH MARBLE AND TRAILSIDE CO-RICH DIOPSIDITE. SAMPLE 3D-2 IS EXCLUDED DUE TO SIGNIFICANT UNDER-REPRESENTATION OF AL IN THE SAMPLE RELATIVE TO THE ROCK UNIT. ............................................................................................................. 201 FIGURE 4.31: IRON CONTENT OF SPINEL COMPARED TO ITS HOST ROCK. SAMPLES INSUFFICIENTLY MINERALOGICALLY REPRESENTATIVE OF ROCK UNIT COMPOSITION WERE EXCLUDED. ONLY THE SPINEL-ROCK PAIR FROM SOPER FALLS HUMITITE DEVIATES FROM THE TREND (EXCLUDED FROM THE LINEAR REGRESSION TRENDLINE). .......................................................................... 202 FIGURE 5.1: PART OF THE WEATHERED LAPIS LAZULI BAND AT THE MAIN SOPER RIVER OCCURRENCE, LOOKING SE TOWARD THE VIOLET SPINEL OCCURRENCE. ................................................................................................................. 207 FIGURE 5.2: LAPIS LAZULI ROCK FROM SOPER RIVER. THE ROCK IS COMPOSED OF PALE YELLOWISH CALCITE, GREY DIOPSIDE, AND BLUE HAÜYNE. ............... 207 FIGURE 6.1: THE PROPOSED GENETIC MODEL FOR LAKE HARBOUR GROUP GEM MINERAL OCCURRENCES. STAGES CONTAINING GEM MINERALS ARE REPRESENTED IN BLUE, AND THOSE DEVOID OF GEM MINERALS IN BLACK.............................................................................................................................................. 220   xxiv  Acknowledgments I wish to thank the following individuals and organizations that provided assistance, knowledge, and support during the course of my PhD.  Dr. Lee Groat (supervisor) always made himself available to discuss plans and share wisdom. Your mentorship has been very important to me. My committee, Drs. James Mortensen and Lori Kennedy, offered useful advice and criticism, as did my committee at the time of candidacy (J. Mortensen, M. Raudsepp, and K. Hickey). My scientific and grant writing skills reached new levels thanks to Mackenzie Parker. Many useful comments were provided by External Examiner Dr. Ian Graham, and University Examiners Drs. Maya Kopylova and Tom Troczynski. I owe a great deal of gratitude to Brad Wilson, who was instrumental in the planning and executing of field work on southern Baffin Island, and with whom I have had many interesting discussions on the subject of Canadian gemstone occurrences. I am very grateful to the staff and management of True North Gems Inc. for granting access to their mining claim, providing data, maps, and for discussions about the geology of the property.  The Qikiqtani Inuit Association (QIA) and staff at the Katannilik Territorial Park allowed access to Inuit-owned and park lands, respectively. I thank the community of Kimmirut for welcoming us to their home, allowing access to municipal lands, and providing much needed help in the field. ᓇᑯᕐᒦᒃ. In the sapphire study (Chapter 3), several co-authors provided useful data and figures (see Preface) which have been used in the present dissertation. Allison Brand provided assistance to Lee Groat in the field.  Andrea Cade helped in sample preparation and data collection for Ar-Ar geochronology.  Ryan Sharpe assisted Mostafa Fayek with boron isotope analysis.  Corundum-xxv  bearing samples from Pitcairn, NY were provided by Dr. George Robinson.  Andrew Fagan, Jan Cempirek, Donald Lake, David Turner provided useful suggestions. Helpful comments from Mackenzie Parker, Canadian Mineralogist guest editor Dr. Dan Marshall, and reviewers Drs. Yannick Branquet and Eloise Gaillou improved the quality of Chapter 3.  In the spinel study (Chapter 4), I wish to thank Holly Steencamp for the Hall Peninsula marble sample, and Glenn Poirier for help with EPMA analysis. Helpful comments from Canadian Mineralogist associate editor Matthew Steele-MacInnis, reviewers D. Skipton, B. Rondeau, B. Dyck, and one anonymous reviewer improved Chapter 4. Financial support was provided by the Dr. Eduard Gübelin Scholarship, a UBC Four-Year Fellowship, UBC tuition awards, a Northern Scientific Training Program in the form of a grant, and by the Natural Sciences and Engineering Research Council of Canada in the form of a Discovery Grant (to Prof. Lee Groat, funding reference 06434) and Canada Graduate Scholarships (CGS-M and CGS-D3). Additional funding was provided by the Mineralogical Association of Canada (Foundation Scholarship), the International Mineralogical Association (Outstanding PhD Student Award), and the Walker Mineralogical Club (Peacock Prize). My time in Vancouver was a lot of fun thanks to my friends DJ Lake, Jaimy Henman, and Ashley Shapiro. I will remember our crazy gem hunting, camping, and snow sculpting adventures forever. I thank Joel Grice, André Lalonde, Michel Picard, John Montgomery, Michael Bainbridge, Rob Woodside, and George & Susan Robinson for their friendship and encouragement of my pursuit of mineralogy. Most importantly, I would not be here were it not for the tireless support of my parents, Michel and Aline, who have always been there for me. 1  Chapter 1. Introduction and Literature Review 1.1 Introduction Sapphire (gem-quality corundum), Mg-Al spinel, and lapis lazuli (a rock principally composed of haüyne, ideally Na3Ca(Si3Al3)O12(SO4) ) are mined for their use in jewelry. Many sapphire deposits, and all in situ spinel and lapis lazuli deposits occur in metacarbonates, but there are few detailed petrologic and geochemical studies on such deposits, especially with regards to sapphire and spinel, for the following reasons: (a) gem deposits are rare, and some varieties of gemstones (e.g., cobalt-blue spinel) are only known from a few localities; (b) many marble-, skarn-, or calc-silicate hosted sapphire and spinel deposits are mined from eluvial deposits formed in part from the weathering of carbonate minerals (e.g., Mogok District, Myanmar; Luc Yen District, Vietnam; Tanzania) while other deposits are alluvial (e.g., Ratnapura, Sri Lanka); (c) many gem deposits are located in politically sensitive areas with limited access (e.g., Myanmar, Afghanistan); and (d) most deposits are mined on an artisanal scale, where no geological records or samples are kept but samples can sometimes be collected in situ during production.  Metacarbonates in the Precambrian Lake Harbour Group (LHG), southern Baffin Island, of high metamorphic grade (granulite and upper amphibolite), are recognized for their outstanding gem potential (Gertzbein 2003). High-quality gem sapphire (LeCheminant et al. 2005, Dzikowski 2013, present research), low-grade lapis lazuli (Hogarth 1971, Hogarth & Griffin 1978), violet spinel, and highly prized vivid cobalt-blue spinel (Wilson 2014) are known to occur in the LHG, especially in the well-explored Kimmirut area.  The gem-bearing rocks of southern Baffin Island offer a unique opportunity to conduct detailed petrological work on a diverse suite of in situ gem occurrences within a single geologic 2  terrane. Such gem occurrences are typically rare and seldom well-exposed. The goal of this study is to develop an understanding of the conditions responsible for the formation of these gem occurrences and to establish deposit models and to formulate exploration strategies. This goal will be achieved by building on existing petrogenetic models together with developing new models for the genesis of gem corundum, spinel, and lapis lazuli in LHG metacarbonates. These models will be constrained by an estimate of protolith composition, the nature of metasomatism, metamorphic history, and trace element geochemistry.  The research comprises three parts: (1) Development of a genetic model for Kimmirut-type sapphire from geochemical, isotopic and petrographic data, and by using thermodynamic modelling to determine the timing and P-T conditions of key mineral assemblage transformations; (2) Development of a genetic model for LHG gem spinel occurrences based on petrology, estimates of the protolith composition (where applicable), and trace element geochemistry (spinel chromophore concentrations) with a focus on determining the possible sources of cobalt-enrichment relative to other chromophores; (3) An evaluation of the lapis lazuli protolith building upon previous research, and measuring Li and other trace element concentrations in lapis lazuli bearing meta-evaporite at Soper River, Baffin Island (Hogarth & Griffin 1978) to provide a basis for the comparison of lapis lazuli meta-evaporite to other metacarbonate-hosted gem occurrences in the LHG.  These new or improved genetic models are expected to facilitate gem exploration in Baffin Island and elsewhere by creating a set of criteria for recognizing areas of high gem potential. In addition, this research will contribute to our general understanding of gem deposits and their relation to regional geology, metamorphic conditions, and protolith geochemistry.  3  1.2 Geology of study area 1.2.1 Regional geology  The sapphire-, spinel-, and haüyne-bearing rocks occur near Kimmirut, in Markham Bay, and on the Hall Peninsula (Fig. 1.1), in a sequence of marble and calc-silicate rocks structurally overlying garnet psammite, meta-semipelite, metapelite, and rare orthoquartzite. The marble, calc-silicate rock, and metaclastic rocks form the Lake Harbour Group (LHG) supracrustal suite (Scott 1997, Scott et al. 1997, 2002). The LHG is interpreted to be the sedimentary cover sequence of the Meta Incognita microcontinent, which was accreted to the southern margin of the Rae Craton by the Trans-Hudson Orogeny (THO; St-Onge et al. 2009). Detrital zircon ages indicate a Paleoproterozoic (rarely Archean) sediment source for LHG sediments and suggest a depositional age for the LHG older than ca. 1.93 Ga (corresponding to the youngest detrital zircon age, Scott et al. 2002). Sparse mafic/ultramafic sills up to 200 m thick and 5.6 km in maximum dimension intrude the LHG and were affected by peak granulite-facies metamorphism (St-Onge et al. 2015). The supracrustal rocks of the THO, including the LHG, were subjected to multiple deformational and metamorphic events (see St-Onge et al. 2007). In deformation event D0, which post-dates the deposition of the LHG and pre-dates emplacement of the Cumberland batholith, basement rocks were imbricated with cover rocks. The Cumberland batholith, an Andean-type granitic batholith, was emplaced from 1865 +4/-2 to 1848 ± 2 Ma. The Narsajuaq oceanic arc was accreted to Meta Incognita (deformation event D1A), and the suture closed between 1845 ± 2 and 1842 +5/-3 Ma. Prograde granulite-facies metamorphism (M1A) occurred ca. 1849 – 1835 Ma as a result of crustal thickening (accretion of the Narsajuaq arc) and heat advection (emplacement of the Cumberland batholith). A high-temperature thermal perturbation 4  (M1B), ca. 1833-1829 Ma, represents either continued mineral growth in the late stages of M1A or a distinct thermal event related to felsic intrusions. Retrograde upper-amphibolite facies metamorphism (M2) occurred from 1820 ± 1 Ma to 1813 ± 2 Ma with reactivation and further shortening of the Soper River suture (D2) which separates Meta Incognita from the Narsajuaq terrane. Recrystallization to S2 assemblages was noted to be most significant in samples with strong foliation and occurring contiguous to D2 thrusts. This localized recrystallization is interpreted by St-Onge et al. (2007) as being driven by deformation-enhanced fluid circulation. Lastly, post-D2 thermal and fluid activity occurred ca. 1797 – 1785 Ma, possibly related to late felsic intrusive rocks of similar age and speculated to be a potential cause of minor greenschist facies retrogression (secondary chlorite, epidote, sericitization). Pressure-temperature determinations for the Lake Harbour Group rocks by St-Onge et al. (2007) are as follows: S1A mineral assemblages provided P-T values of ca. 810 °C and 8.0 kbar; thermobarometry of S2 mineral assemblages indicate metamorphic conditions of upper amphibolite facies, ca. 720 °C and 6.2 kbar. LHG rocks, including marbles and calc-silicates, in the western part of the Hall Peninsula were metamorphosed to granulite facies (~850-890 °C, 6.1-7.35 kbar) during 1850-1825 Ma (Skipton et al. 2016). 1.2.2 Lake Harbour Group marble and calc-silicate rocks Marbles and calc-silicate schists showing centimeter- to meter-scale compositional layering comprise a significant portion of the Lake Harbour Group (Scott 1997). The LHG metacarbonate and calc-silicate rocks commonly contain calcite, diopside, phlogopite, forsterite, serpentine, amphibole, spinel, graphite, pyrite, pyrrhotite, or scapolite with the occasional occurrence of apatite, titanite, wollastonite, humite, plagioclase, or magnetite (Blackadar 1967, Scott 1997, Herd et al. 2000, St-Onge et al. 2000). Grain sizes range from fine to very coarse. 5  Nepheline and K-feldspar occur in calc-silicate bands and marble on Aliguq Island, 165 km NW of Kimmirut (Butler 2007). Layering in the marbles is generally convoluted, probably due to ductile flow (Scott 1997), and silicate-rich lithologies occur as both stratigraphically continuous layers and boudins or pods (Butler 2007). 1.3 Broad research questions and significance Some high-grade, metacarbonate-rich terranes contain multiple gem spinel (Hughes 1997), ruby (Garnier et al. 2008 and references therein), sapphire (Hughes 1997), and uncommon lapis lazuli deposits, such as that in Vietnam (Chauviré et al. 2015, Pham et al. 2013, Blauwet 2006), Afghanistan (Faryad 2002), Tajikistan (Hughes & Pardieu 2006), and Myanmar (Pardieu 2014, Themelis 2008). Other terranes of similar metamorphic grade and abundance of metacarbonates are not necessarily known to contain any significant gemstone occurrences. Such an example is the Central Metasedimentary Belt (CMB, Grenville Province, Ontario-Quebec; Rivers 2015, Corriveau 2013), from which noteworthy coloured gemstone occurrences are not yet known, but the CMB is known to contain spinel and corundum of poor colour and quality (e.g., Wilson 2014, Belley et al. 2016) in addition to gem-quality skarn-hosted grossular (Belley & Bourdeau 2014). The metacarbonate rocks of the Lake Harbour Group, Baffin Island contain numerous and significant gemstone occurrences which are of particular significance due to the size and quality of gemstones (sapphire up to 7.81 carats), the occurrence of rare coloured stones (vibrant cobalt-blue spinel), and in some cases, the extent of mineralization (low-grade lapis lazuli/haüyne). These well-exposed gem occurrences offer a unique opportunity for research: A comprehensive study of the genesis of metacarbonate-hosted gemstones of diverse mineralogy in a single geologic terrane. The present dissertation seeks to determine the ways in which regional 6  geologic phenomena influence gemstone potential (e.g., metamorphic history, metasomatism, intrusions, and protoliths) by means of creating genetic models for gem corundum, spinel, and lapis lazuli in the LHG.  An understanding of these regional-scale controls on gem formation in high-grade metasedimentary belts helps clarify the origin of coloured stones that have fascinated human societies for millennia and facilitate exploration. The present research contributes new information on the effects of metamorphic history, protolith composition, and chromophore distribution on gemstone genesis which enhances our general understanding of metamorphic gem formation. 1.3.1 Part 1 - Corundum Calc-silicate-hosted sapphire occurrences near Kimmirut, Baffin Island, comprise a new type of gem corundum deposit (see section 1.4.1 for a review of known deposit types). Developing an understanding of corundum genesis in these scapolite-rich rocks represents an important new contribution to the understanding of gem-quality sapphire formation and helps predict the potential for such deposits in the LHG and other high-grade marble-rich terranes. Kimmirut sapphire was previously studied by LeCheminant et al. (2005) and Dzikowski (2013), but their early interpretations did not fully explain the origin of corundum in these calc-silicate pods. LeCheminant et al. (2005) proposed that the calc-silicate pods formed from contamination of a syenitic intrusion and were subsequently metamorphosed. Dzikowski (2013) contradicted the previous study, suggesting a mixed evaporite-black shale protolith, where the corundum-bearing assemblage was formed during retrograde metamorphism from the possible break-down of nepheline + scapolite, nepheline + anorthite, or anorthite. 7  The goal of the present research is to produce a robust deposit model for Kimmirut sapphire using a combination of new data and a different approach to solving the complex petrogenesis of sapphire-bearing calc-silicate rock using additional mineralogical and textural observations, petrography, X-ray powder diffraction, and boron isotope geochemistry. The approach to protolith determination used in the present study yielded different results from previous workers. The geochronology of mineralization (U-Pb of zircon, 40Ar/39Ar of micas) was re-interpreted, yielding different results to previous authors. 1.3.2 Part 2 - Spinel Several areas in the world are producing gem spinel, and much of this spinel is recovered from eluvial or alluvial deposits. Interest in spinel gemstones has increased significantly in the early 21st century (Pardieu et al. 2008); however, the geology of spinel has yet to receive attention from geoscientists – in significant contrast to gem corundum geology (e.g., Giuliani et al. 2014). Giuliani et al. (2017) noted that “[g]eological investigations of gem spinel deposits in marble are scarce” and that “[s]everal questions remain regarding their genesis, particularly: (1) their age and the P-T-x conditions of formation; (2) the relationship between oxygen isotopic composition of spinel and the nature of fluid-rock interaction with marbles; and (3) the sources of Al and other trace elements incorporated in spinel.” The potential for gaining new insight into gem spinel genesis from small in situ occurrences that are not being mined for gem material (giving a much larger sample size) has been overlooked. The present dissertation presents the first comprehensive gem spinel deposit model, which is based on fourteen occurrences in the Lake Harbour Group, including two cobalt-blue spinel occurrences, and seeks to determine geochemical controls on whether or not spinel forms in a metacarbonate rock. 8  Spinel of an intense red colour (chromium chromophore) and vibrant cobalt-blue colour (cobalt chromophore) are of greatest commercial interest. Ruby (Cr-bearing corundum) deposit models suggest that Cr enrichment is a result of Cr-rich sediment deposited coevally with carbonates and the concentration of Cr in organic matter (Garnier et al. 2008). At Revelstoke, British Columbia, Cr was enriched in gneiss layers that reacted with marble to form corundum (Dzikowski et al. 2014). The enrichment of cobalt in concentrations sufficient to turn spinel bright blue has yet to be adequately explained. In the most detailed petrologic study on cobalt-blue spinel to date, Chauviré et al. (2015) suggested that Co and Ni were enriched in cobalt-blue spinel from F- and Cl-bearing evaporite-contaminated metamorphic fluids sourcing these metals from evaporite or amphibolite. This hypothesis will be evaluated for the Kimmirut occurrences. The effect on chromophore element distribution in minerals associated with spinel, including preferential incorporation of elements in some phases over others, has yet to be studied or assessed as a potential control on the colour of gem spinel. For example, does an abundance of sulphide minerals (e.g., pyrite, pyrrhotite) preferentially incorporate Co, thus limiting its availability to spinel? Or is the effect negligible at Co concentrations in typical cobalt-blue spinel-containing rocks? 1.3.3 Part 3 - Lapis lazuli  Meta-evaporite compositions are diverse (e.g., Moine et al. 1981), and major element rock compositions can overlap with normal sediments. Concentrations of trace elements, most notably Li, offer good potential for meta-evaporite identification, however Li tends to be poor in dolomite-rich evaporite-related rocks (Moine et al. 1981). The trace element composition of lapis lazuli at Soper River will be used as a check primarily for Li concentrations in an example of Precambrian meta-evaporite in the LHG. Furthermore, the protolith composition of lapis lazuli 9  will be estimated based on whole rock chemical composition to establish differences and similarities with the composition of other metacarbonates containing gem minerals. 1.4 Background on gem corundum 1.4.1 Geology of gem corundum deposits  Gem corundum (sapphire: blue, purple, pink, colourless, yellow, etc.; ruby: red) occurs in a variety of geological environments where Al is in excess relative to Si (i.e., Si is a limiting reactant in aluminosilicate formation), and where P-T-x conditions favor corundum stability (Giuliani et al. 2014). Gem corundum occurs in igneous rocks (syenite, monzonite, xenocrysts in basalt and lamprophyre), metasomatic rocks (skarn, biotitite, desilicated pegmatite and gneiss), metamorphic rocks (marble, mafic granulite, amphibolite, aluminous gneiss), and anatexite (Fig. 1.2, Giuliani et al. 2014, Simonet et al. 2008). Since this research pertains to metacarbonate-related gem deposits specifically, these are described in greater detail below.  Metamorphic gem corundum deposits form under amphibolite and granulite facies conditions at 500 to 800 °C temperature and 2 to 10 kbar pressure (Fig. 1.3, Giuliani et al. 2014). Four types of metacarbonate-related deposits have been defined (see reviews by Giuliani et al. 2014 and Simonet et al. 2008): (1) Endoskarns where magma intrudes marble and becomes desilicated, such as the pargasite-phlogopite-hosted ruby occurrence at the edge of mafic dykes at Mahenge, Tanzania; (2) Exoskarns where pegmatite or granite intrudes marble or calc-silicate rocks, such as in Myanmar (Themelis 2008), Sri Lanka (scapolite-corundum-spinel zoned skarn, Silva & Siriwardena 1988), and Madagascar (diopsidite and scapolitite/marble); (3) Stratiform gem corundum mineralization in calcite-dolomite marble, typically formed from the retrograde break-down of spinel, and at some localities suggested to be metamorphosed evaporite lenses 10  (Garnier et al. 2008); (4) Impure marble containing gneiss or silicate-rich layers, which react with the marble under peak metamorphic conditions (e.g., Dzikowski et al. 2014). 1.4.2 Colour of corundum  Pure corundum (Al2O3) is colourless, and colourless sapphire gemstones are less sought after than yellow, blue, and red, in order of increasing value. Yellow colouration in corundum is caused by O2-Fe3+ charge transfer. Corundum is made blue by the presence of Fe and Ti, which cause the colour by intervalence charge transfer (Fe2+Ti4+; Fritsch & Rossman 1988). Ti4+ preferentially pairs with Mg2+ relative to Fe2+, and Mg2+ preferentially pairs with Si4+ relative to Ti4+; therefore, Si and Mg trace concentrations control the availability of Ti4+ to Fe2+, and thus influence the colour of corundum (Emmett et al. 2017).  Red hues (i.e., ruby or pink sapphire) are caused by the presence of Cr3+ (Fritsch & Rossman 1987). 1.4.3 Previous studies on Kimmirut sapphire deposits LeCheminant et al. (2005) suggested that formation of the corundum-bearing calc-silicate rocks at the Beluga and Bowhead occurrences involved a contaminated syenitic magma emplaced in marble late in the D2 deformation event.  The nepheline- and diopside-rich assemblage was proposed to have been formed by the reaction of syenitic magma with LHG marble and possibly evaporites, followed by retrogression forming phlogopite-oligoclase-scapolite at the expense of diopside and nepheline, and a second retrograde event resulting in the fracture-controlled alteration of nepheline to form corundum.  In a study of nepheline-bearing LHG metacarbonates on Aliguq Island, Butler (2007) determined that nepheline-K-feldspar formed late based on textural relations with forsterite, diopside, and phlogopite. Butler (2007) proposed that this assemblage formed at peak metamorphic conditions (ca. 800 °C and 8 kbar) at low XCO2 (≤ 0.15) by the reaction 3 albite + 3 calcite + phlogopite  K-feldspar + 3 nepheline + 11  3 diopside + H2O + 3 CO2.  Dzikowski (2013) suggested that the protolith for the Beluga and Bowhead calc-silicate pods was evaporite-black shale deposited in the LHG carbonate shelf and proposed the following paragenetic sequence: (1) prograde diopside + nepheline (≤ 810 °C, 8.3 kbar); (2) alteration of the peak assemblage by NaCl-bearing fluids (at < 710 °C, 6 kbar) forming phlogopite-oligoclase ‘symplectites’ with late scapolite rims; (3) introduction of hydrous fluids, post-D2, causing the break-down of nepheline and either scapolite or anorthite to form albite, muscovite, and corundum or alternatively, the introduction of Na-bearing hydrous fluids breaking down anorthite to albite and corundum. Lepage & Davison (2007) reported the use of ultraviolet LED technology to explore for fluorescent scapolite, which they assumed would coexist with gem corundum in the Kimmirut area.  Turner et al. (2017) investigated the potential application of hyperspectral imaging in exploration for Kimmirut-type gem corundum deposits. In Chapter 3, this dissertation re-evaluates the data presented by Dzikowski (2013) with additional isotopic and petrographic data, and applies the results to change, refine, and generate new interpretations on the origin and petrogenesis of gem corundum in Kimmirut-type deposits. 1.5 Background on gem spinel 1.5.1 Geology of gem spinel deposits Mg-Al spinel is a common constituent of marble and metasomatic rocks associated with marble, however it is seldom of gem quality due to its typically unattractive dark colour (e.g., Van Velthuizen 1993, Belley et al. 2016) and small crystal size. Gem-quality Mg-Al spinel, which is most prized when intense red, pinkish-red, or cobalt-blue in color, is mined at several localities worldwide, most of which occur in marble metamorphosed at high metamorphic grade: (1) red and pinkish-red spinel is mined in marble and eluvial deposits in karsts at Mogok and 12  Namya, Myanmar (Pardieu 2014, Themelis 2008); (2) pinkish red spinel, the largest crystal having been found to date weighing 54 kg and the largest gemstone weighing 50 carats, is mined from marble and eluvial deposits near Ipanko, Tanzania (Pardieu 2008); (3) red, pink, blue, and cobalt-blue spinel is mined from granulite-facies marble and contiguous eluvial and alluvial deposits in the Luc Yen District, Vietnam (Chauviré et al. 2015, Pham et al. 2013, Senoble 2010, Blauwet 2006); (4) blue, purplish-blue, and greyish cobalt-blue spinel is mined from alluvial placer deposits in Ratnapura, Sri Lanka (Shigley & Stockton 1984, personal communication M. Ikram); and (5) red and pink gem-quality spinel occurs in marble and magnesian skarns (at the contact of gneiss and marble) at Kuh-i-Lal, Tajikistan (Kievlenko 2003, Hughes & Pardieu 2006). Giuliani et al. (2017) identified several questions that remain regarding gem spinel genesis, notably about the P-T-x conditions of formation and the sources of Al and trace elements incorporated in spinel. Most of these questions have been addressed in the present dissertation, with the notable exception of Cr-bearing red spinel, which has not been found in the Lake Harbour Group on Baffin Island. 1.5.2 Colour of spinel  Spinel color is determined by the concentration of Fe, Cr, Co, and to a lesser degree, V, in addition to the coordination and charge of Fe (D’Ippolito et al. 2015). As total Fe content increases in spinel, it grades in color from colorless to pale lilac (e.g., 0.58-0.65 wt. % FeOtotal Fe, Kleišmantas & Daukšytė 2016), sky blue, green, deep green (Hålenius et al. 2002) and black (Van Velthuizen 1993). Pink and red spinel is primarily colored by Cr3+ in concentrations of at least 0.1 wt. % Cr2O3, and color saturation is largely determined by Fe content (Kleišmantas & Daukšytė 2016). Kleišmantas & Daukšytė (2016) suggested that V contributes yellow or brown 13  hues to spinel in concentrations over 0.4 wt. % V2O3. It should be noted that the study of Kleišmantas & Daukšytė (2016) compared spinel chemical compositions directly with colour without supporting the observed colour-composition trends using absorption spectra. Cobalt-poor spinels containing Fe are typically blue (octahedral Fe3+), purplish-blue (tetrahedral Fe2+), or greenish blue (higher total Fe), and TCo2+ absorption bands can be detected in spinel at Co concentrations greater than 10 µg/g (D’Ippolito et al. 2015). Natural spinel with a significant Co color contribution was first noted by Shigley & Stockton (1984). Vibrant cobalt-blue spinel from Vietnam has a Fe/Co ratio of 10, a Co concentration of 1236 µg/g, ~2500 µg/g Ni, ~1100 µg/g Cr, and ~1000 µg/g Zn. The Cr content of this spinel gives it a very slight purplish tinge when viewed in incandescent light (Chauviré et al. 2015). Based on trace element and light absorption data, Chauviré et al. (2015) estimated that the visible light molar absorptivity of cobalt in spinel is ~20× more important than that of iron. 1.5.3 Previous studies on Lake Harbour Group spinel Spinel occurrences in Nunavut have not yet been the subject of scientific research. Wilson (2014) noted spinel from the following localities within the LHG: (1) violet crystals up to 3 cm in marble near the Soper River lapis lazuli occurrence near Kimmirut; (2) blue and violet spinel crystals up to 3 cm at Waddell Bay; (3) violet crystals up to 4.5 cm in a mica exploration trench on Soper Lake, near Kimmirut; (4) violet crystals in ‘chondrodite’ at Soper Falls, north of Kimmirut; (5) violet crystals up to 5 cm at Glencoe Island (see also Grice et al. 1982); (6) cobalt-blue, opaque fragments up to 1 cm at the Qila occurrence, near Kimmirut; and (7) cobalt-blue, occasionally transparent and heavily fractured crystals up to 2.7 cm at the Trailside occurrence, also near Kimmirut. Small faceted gemstones, almost all under 0.50 carats (with the exception of 14  an opaque 2.14 carat Qila spinel cabochon) have been produced from these localities (Wilson 2014). 1.6 Background on lapis lazuli 1.6.1 Lazurite vs. haüyne Lapis lazuli is commonly reported to be composed of the sodalite group minerals lazurite and/or haüyne. Lazurite is the sulphide end-member, ideally Na6Ca2(Al6Si6O24)S2, and haüyne is the sulphate end-member, Na3Ca(Al4Si4O12)(SO4). Multiple studies indicate the occurrence of lazurite, however in examples where sulphate and sulphide content was determined in the mineral, the sulphate member rather than sulphide is the dominant species, including at Sar-e-Sang, Afghanistan and Soper River, Baffin Island, Canada, (Hassan et al. 1985, Fleet et al. 2005, Moore & Woodside 2014) with some examples being sulphite-dominant (Tauson et al. 2012). Therefore, natural examples of “lazurite” are in fact all haüyne. 1.6.2 Geology of lapis lazuli deposits Lapis lazuli is a royal blue rock composed primarily of haüyne and is used as an ornamental stone and in jewelry. Three economic lapis lazuli deposits (Afghanistan, Russia and Chile) are well-known known, as are a number of small deposits or deposits where the rock is too impure to be of commercial interest.  At Coquimbo, Chile, lapis lazuli composed primarily of haüyne, wollastonite, calcite, diopside, pyrite, and scapolite occurs in lenses up to 2 × 0.1 m (locally up to 0.4 m in thickness) in three skarn zones up to 300 m wide. The deposit formed from the contact metamorphism and metasomatism of limestone by granite (Coenrads & Canut de Bon 2000). Smaller occurrences of lapis lazuli formed by contact metamorphism and metasomatism include Cascade Canyon in California, USA (quartzite/limestone protolith; Housley 2012 and 15  references therein), Latium, Italy (limestone xenoliths in leucitic tuff, Hogarth & Griffin 1975), and at Italian Mountain, Colorado, USA (subgreywacke/dolostone protolith; Hogarth & Griffin 1980).   Two deposits, Tultui and Malo Bystrinsky, occur near Lake Baikal, Russia (Hogarth 1979). An unspecified deposit near Lake Baikal is reported to have formed by contact metasomatism of marble by alkaline igneous rocks (Deer et al. 2004 and references therein).  The lapis lazuli deposit at Sar-e-Sang, Afghanistan was regionally metamorphosed, possibly subjected to metasomatism, and the lapis lazuli is found in three distinct modes of occurrence: (1) At the contact with granite and pegmatite, lenses up to 40 × 2 m with K-feldspar-quartz-plagioclase cores surrounded by calc-silicate rock composed of diopside, forsterite, phlogopite, tremolite, grossular, etc., and an exterior rim of calcite, diopside, haüyne, nepheline, and afghanite. (2) Lenses up to 450 × 6 m consisting of an alternating sequence of calcite-diopside-haüyne, diopside, and phlogopite-diopside calc-silicate rocks thought to represent original bedding. (3) Lenses of haüyne up to 2 × 0.3 m occurring between granite or pegmatite and marble (Faryad 2002 and references therein). The Sar-e-Sang deposit is considered to be a meta-evaporite contained within a metamorphosed shallow marine sediment sequence (Faryad 2002).  At the Edwards Zn-Pb mine, Edwards, New York, two pods of lapis lazuli up to 2.4 × 0.9 m were found during mining. The core of the pods is composed of haüyne, calcite, actinolite, diopside, and pyrite and is surrounded by a pyrite-rich actinolite layer, a phlogopite layer, and at the contact with dolomitic marble, high-grade sphalerite (Lessing & Grout 1971). A cobble of impure lapis lazuli (haüyne in diopsidite) was found in gravel from glacial till in Ottawa, 16  Ontario, and likely originates from the nearby Central Metasedimentary Belt (unpublished data, P.M. Belley; personal communication Michael Bainbridge).  1.6.3 Geology of the Soper River lapis lazuli occurrence  Along the Soper River, north of Kimmirut, Nunavut, impure lapis lazuli (North and Main occurrences) occurs in five zones (four of which are located at the Main occurrence), up to 168 m long, interbedded with marble. The lapis lazuli rock is composed of varying concentrations of haüyne (up to 42.2 vol. %), diopside, plagioclase, and calcite with locally abundant phlogopite, nepheline, tremolite, and scapolite, and common subordinate pyrite (Hogarth 1971). It is proposed to be a regionally metamorphosed evaporite-limestone (Hogarth 1971, Hogarth & Griffin 1978).  Hogarth & Griffin (1978) convincingly proposed that the Soper River lapis lazuli is a meta-evaporite citing the following evidence: (1) The lapis lazuli has well-developed layering parallel to the regional foliation, suggesting that it is metasedimentary; (2) the area has a scarcity of intrusive rocks, which does not support a contact metasomatic origin; and (3) the abundances of Na, K, S, Cl, Br, F, and Fe are consistent with evaporite-related sediments. 17  1.7 Figures   Figure 1.1: Geology of southern Baffin Island and location of the study areas. Crustal sutures separating structural domains are represented by dashed lines. BS: Bergeron suture; SRS: Soper River suture. Modified after St-Onge et al. (2000) and Butler (2007); ages from St-Onge et al. (2001).  18   Figure 1.2: Classification of gem corundum deposits (after Simonet et al. 2008).  Figure 1.3: P-T conditions for the formation of corundum in various metamorphic deposits (after Giuliani et al. 2014).  19  Chapter 2. Methods 2.1 Sapphire 2.1.1 Petrography and Mineral Identification Approximately 25 kg of representative rock samples were collected by Lee Groat, Paul Gertzbein and others at the Beluga sapphire occurrence and to a lesser degree at the nearby Beluga occurrence (calc-silicate with no sapphire). Samples were studied in the petrological microscope and scanning electron microscope (SEM) from ~16 thin-sections (3 Bowhead). Mineral phases that were not analyzed with electron probe microanalysis (EPMA) were identified using energy dispersive spectroscopy (EDS) in the SEM, or standard X-ray powder diffraction (XRPD) techniques (noted in text). 2.1.2 Chemical analysis Chemical analyses for all minerals with the exception of tourmaline were performed by Dzikowski (2013) with a CAMECA SX-50 electron microprobe (at the University of British Columbia) in wavelength-dispersion (WD) mode. Full data sets are in appendices of Dzikowski (2013).  The operating voltage was 15 kV, with 20 nA beam current and 5 µm beam diameter.  Counts were collected for 20 s for each element with the exception of F and Cl (50 s) and La, Ce, Pr, Nd, Sm, Gd, Th, and U in zircon (60 s).  The following standards were used (all Kα lines): synthetic phlogopite (F), albite (Na), synthetic phlogopite (Mg and Si in phlogopite, K in micas), diopside (Mg, minerals other than phlogopite), kyanite (Al in diopside and micas), corundum (Al in corundum), anorthite (Al in other minerals), zircon (Si in zircon), barite (S), scapolite (Cl), orthoclase (K in micas, Si in muscovite, nepheline, muscovite, and scapolite), rutile (Ti), synthetic V (V), synthetic magnesiochromite (Cr), synthetic rhodonite (Mn), and synthetic fayalite (Fe).  The following additional standards were used for zircon (all Lα lines except UMα): 20  Y3Al5O12 (Y), zircon (Zr), Ca-Al-Si-glass (Drake & Weill 1972; La, Pr, Nd, Sm, Gd), CeO2 (Ce), ThO2 glass (Th), and UO2 glass (U).  Matrix correction calculations were done with using the 'PAP' (Z) method (Pouchou & Pichoir 1985). Chemical analysis of tourmaline was performed with a CAMECA SX100 instrument at Masaryk University, Brno, Czech Republic.  The operating voltage was 15 kV with 10 nA beam current and 5 µm beam diameter.  Counts were collected for 10 s with the following exceptions: 15 s (V), 20 s (Mg, Cr, Ca, Zn), 30 s (Cl), 40 s (K), and 60 s (F).  The following standards were used (all Kα lines): topaz (F), albite (Na), pyrope (Mg), sanidine (Al, Si, K), fluorapatite (P), vanadinite (Cl), wollastonite (Ca), titanite (Ti), scandium vanadate (V), chromite (Cr), spessartine (Mn), almandine (Fe), and gahnite (Zn).  B2O3 concentrations of 10.63 wt. % were estimated for matrix correction of tourmaline data.  Matrix correction was done using the X-PHI method (Merlet 1994). 2.1.3 Boron isotopes The boron isotope composition of tourmaline was determined at the University of Manitoba with a Cameca IMS 7f ion microprobe using secondary-ion mass spectroscopy (SIMS), a primary O- beam (5 nA accelerated at 12.5 kV), with a 15 µm beam diameter, a sample accelerating voltage +10kV, electrostatic analyzer +10 kV, and ETP 133H electron multiplier coupled with an ion counting system having an overall deadtime of 21 ns.  The entrance slit was set at 36 µm with a mass resolving power of 1450.  Counts for 11B, 10B, and 30Si were collected in succession for 50 cycles with 1 s measurements of each isotope per cycle, 30 s pre-sputter, and 0 V offset.  The analytical procedure was similar to that used by Chaussidon & Albarède (1992).  Instrumental mass fractionation (IMF) and analytical quality were assessed by replicate analyses of an elbaite tourmaline reference material (No. 98114, see below); 21  repeatability of the reference material was 0.4‰. Precision (1σ) for the unknown sample during the sessions was ±0.3‰.  The boron isotope composition is expressed in delta notation, as a per mil deviation from boric acid standard NIST RM 951 (11B/10B = 4.0437 ± 0.0033, Catanzaro et al. 1970): δ11B = ([11B/10B]sample/[11B/10B]SRM951 – 1)*1000.  For the analysis of Beluga oxy-dravite, elbaite standard No. 98114 (11B/10B = 4.0014 ± 0.0007, Leeman & Tonarini 2001) was used.  Ludwig et al. (2011) reported no significant matrix effects for B isotope analysis of tourmaline (using dravite, elbaite, and schorl standards) with SIMS.  Cabral et al. (2012) found a 1‰ to 2‰ IMF offset with elbaite, comparable to overall uncertainty and therefore not significant.  However, a +1.6‰ discrepancy in Leeman & Tonarini (2001) standard elbaite No. 98114 relative to dravite No. 108796 was deemed significant by MacGregor et al. (2013). 2.1.4 Oxygen isotopes of corundum The oxygen isotope composition (18O/16O) of corundum was measured at the Isotope Geosciences Unit, Scottish Universities Environmental Research Centre, Glasgow, Scotland using the laser fluorination method of Giuliani et al. (2005).  Precision (1σ) of analyses on a quartz standard is ±0.1‰.  Data is reported in delta notation as a per mil deviation from the 18O/16O value of Vienna Standard Mean Ocean Water (VSMOW) standard NIST RM 8535. 2.1.5 Whole-rock geochemistry Whole rock major and trace elements were analyzed by ALS Chemex in Vancouver, Canada using a combination of inductively coupled plasma mass spectrometry (ICP-MS) and atomic emission spectroscopy (ICP-AES).  Carbon was determined by combustion furnace, and samples were subjected to a lithium borate fusion for resistive elements, a four acid digestion, and aqua regia digestion. Precision of major element analyses, carbon, sulphur and chlorine is 22  5%. Precision of trace element, boron, and fluorine analyses is 10% with the exception of mercury (15%) and copper (7%). 2.1.6 Radiometric dating Zircon crystals were analyzed using conventional ID-TIMS (isotope dilution thermal ionization mass spectrometry) at the Pacific Center for Isotopic and Geochemical Research (PCIGR), University of British Columbia, using the methods described by Mortensen et al. (1995) and Beranek & Mortensen (2011).  Errors attached to individual analyses were calculated using the numerical error propagation method of Roddick (1987), and decay constants are those recommended by Steiger & Jäger (1977).  Compositions for initial common Pb were taken from the model of Stacey & Kramer (1975).  The zircon grains were strongly air abraded prior to dissolution in order to try to minimize the effects of post-crystallization Pb loss. Muscovite and phlogopite were analyzed for 40Ar/39Ar dating in the Noble Gas Laboratory of the PCIGR.  Mineral separates were hand-picked, washed in acetone, dried, wrapped in aluminum foil, and stacked in an irradiation capsule with a neutron flux monitor (Fish Canyon Tuff sanidine, 28.02 Ma, Renne et al. 1998) and 1071 Ma hornblende HB3Gr as an age check (which yielded a flat J-curve and an age of 1069 ± 2 Ma).  The samples were irradiated at the McMaster Nuclear Reactor in Hamilton, Ontario, for 90 MWH, with a neutron flux of approximately 6 × 1013 neutrons/cm2/s.  Analyses (n = 45) of 15 neutron flux monitor positions produced errors of <0.5% in the J value.  At PCIGR, the mineral separates were step-heated at incrementally higher powers in the defocused beam of a 10W CO2 laser (New Wave Research MIR10) until fused.  The gas evolved from each step was analyzed with a VG5400 mass spectrometer equipped with an ion-counting electron multiplier. All measurements were corrected for total system blank, mass spectrometer sensitivity, mass discrimination, radioactive 23  decay during and subsequent to irradiation, atmospheric Ar contamination, and the effect of irradiation on Ca, Cl, and K.  The plateau and correlation ages were calculated using Isoplot v3.09 (Ludwig 2003).  Errors are quoted at the 2-sigma (95% confidence) level and are propagated from all sources except mass spectrometer sensitivity and age of the flux monitor.  The best statistically justified plateau and plateau age were picked based on the following criteria: (1) three or more contiguous steps comprising more than 50% of the 39Ar; (2) probability of fit of the weighted mean age greater than 5%; (3) slope of the error-weighted line through the plateau ages equals zero at 5% confidence; (4) ages of the two outermost steps on a plateau are not significantly different from the weighted-mean plateau age (at 1.8σ, six or more steps only); and (5), the outermost two steps on either side of a plateau must not have nonzero slopes with the same sign (at 1.8σ, nine or more steps only). 2.2 Spinel 2.2.1 Sampling, petrography and mineral identification The sampling strategy used in the field on southern Baffin Island was to extract samples from each unique lithology at a spinel occurrence (both spinel-bearing and spinel-absent rocks, with as representative a sample as possible) and, where possible, to sample nearby marble, calc-silicate, and other rocks for geochemical comparisons. Rock samples were examined in thin-section using an petrological microscope and scanning electron microscope (SEM). Mineral phases that were not analyzed with electron probe microanalysis (EPMA) were identified using energy dispersive spectroscopy (EDS) in the SEM, or standard X-ray powder diffraction (XRPD) techniques (noted in text).  24  2.2.2 Electron probe microanalysis Minerals analyzed with EPMA were extracted using hand tools directly from the rock samples (of which parts were used for whole rock geochemical analysis), mounted in epoxy pucks, and polished. Chemical compositions were obtained with a JEOL JXA-8230 electron microprobe (University of Ottawa) in wavelength-dispersion (WD) mode. The operating voltage was 20 kV, with 20 nA beam current and 5 µm beam diameter. Counts were collected for 20 seconds for each element with the exception of F (30 seconds). The following standards were used (all Kɑ except for Ba Lɑ): Sanidine (Si, Al, K, Ba), albite (Na), diopside (Ca, Mg), dolomite (Ca and Mg in carbonates), hematite (Fe), olivine (Fe, Mg and Si in forsterite), pyrrhotite (Fe and S in sulphides), chromite (Cr, Co, Al in forsterite and spinel, Mg in spinel), rutile (Ti), tephroite (Mn), pentlandite (Ni), vanadinite (V), gahnite (Zn), tugtupite (Cl), and fluorite (F). Matrix correction calculations were done using the Armstrong/Love-Scott ϕ(ρZ) method (Armstrong 1988). 2.2.3 Whole-rock geochemistry Whole rock major and trace elements were analyzed by ALS Canada Ltd in Vancouver, Canada. Samples were crushed and pulverized. Major elements (Si, Al, Fe, Ca, Mg, Na, K, Ti, Mn, in addition to Ti, P, Sr, Ba) were determined by inductively coupled plasma – atomic emission spectroscopy (ICP-AES) following fused bead preparation and acid digestion. Loss on ignition was determined by heating samples in a furnace at 1000 °C. Total sulphur and carbon were determined by Leco furnace. The following elements were measured by inductively coupled plasma – mass spectrometry analysis (ICP-MS) following Li borate fusion and acid digestion: Ba, Sr, Cs, Cr, V, Y, REE, Hf, Zr, Nb, Ta, Th, U, W, Rb, Ga, Ge, and Sn. The elements As, Bi, Hg, In, Re, Sb, Se, and Tl were determined by ICP-MS. The following elements 25  were measured by ICP-AES following Li borate fusion and four acid digestion: Ag, Cd, Co, Cu, Li, Mo, Ni, Pb, Sc, and Zn. Boron was determined for selected samples using ICP-MS following NaOH fusion. Chlorine and F were determined for selected samples using ion chromatography following KOH fusion. Precision of major element analyses, carbon, sulphur and chlorine is 5%. Precision of trace element, boron, and fluorine analyses is 10% with the exception of mercury (15%) and copper (7%). 2.2.4 Thermodynamic modelling Pseudo-sections were generated with Perple_X version 6.8.1 (Connolly 2009) using the default thermodynamic database (hp02ver.dat), the default computational file, and the CORK fluid equation of state (Holland & Powell 1991, 1998), with calculations assuming a saturated CO2-H2O fluid, using a constrained 2d grid computational model, and the following solution models: GlTrTsPg for clinoamphiboles; Sp(GS) for spinel; O(SG) for forsterite; Bio(HP) for phlogopite; Cpx(l) for diopside; and Scap for scapolite. A humite-rich sample could not be modeled due to the lack of humite in the thermodynamic databases (clinohumite is present, but is structurally and chemically different to humite). The accuracy and precision of calculated phase diagrams is sensitive to uncertainty related to petrographic variations at hand sample and thin-section scales, e.g., displacement of equilibria by ±1 kbar for a moderate degree of modal proportion uncertainty (20% relative threshold, Palin et al. 2016). Spinel-bearing rocks on Baffin Island tend to be very coarse-grained, and heterogeneous in texture and mineral abundance. Therefore, only highly representative samples were selected for thermodynamic modelling (samples 2A-SPL-2, 3A-1, 3E-3-A, and 3F-2). The nature of the spinel occurrences and the limited volume of sampling may nonetheless be a source for error, since the chosen samples may not perfectly represent the total 26  equilibration volume. Sample 3F-1 was excluded due to the high abundance of sericitization, which may have affected the bulk composition. Bulk rock compositions used in the calculations were obtained from whole rock analysis of crushed 0.3-0.6 kg hand samples. Whole rock concentrations of Fe (sample 2A-SPL-2 only) and Ca were corrected to compensate for the presence of pyrrhotite and apatite, respectively, which have been excluded from the effective bulk composition. 2.3 Lapis lazuli 2.3.1 Whole-rock geochemistry  Two whole rock Soper River lapis lazuli samples were analyzed using the same method as the spinel study.   27  Chapter 3. Origin of scapolite-hosted sapphire (corundum) near Kimmirut 3.1 Results summary Gem-quality corundum (sapphire) occurs in scapolite-rich calc-silicate rock hosted in marble of the Lake Harbour Group near Kimmirut, southern Baffin Island. A deposit of blue and colourless gem corundum (Beluga occurrence) is compared to a similar calc-silicate pod generally lacking corundum but containing nepheline (Bowhead occurrence) and located 170 m to the SSW. Corundum formation was made possible by three equally important sequential metamorphic reactions: (1) formation of nepheline, diopside, and K-feldspar (inferred) at granulite facies peak metamorphic conditions; (2) partial retrograde replacement of the peak assemblage by phlogopite, oligoclase, calcite, and scapolite (Me50-Me67) as a result of CO2-, H2O-, Cl-, F-bearing fluid influx at 1782.5 ± 3.7 Ma (P-T < 720 °C, 6.2 kbar); and (3) retrograde break-down of scapolite + nepheline (with CO2- and H2O-bearing fluid) to form albite, muscovite, corundum, and calcite. Late, low-temperature zeolite mineralization is common in corundum-bearing zones. Based on thermodynamic models, the corundum-forming reaction only occurs in a <100 °C window with an upper limit determined by scapolite-nepheline stability, and a lower limit determined by the formation of Al-silicate rather than corundum.  The protolith is inferred to be dolomitic argillaceous marl with no evidence to suggest the initial presence of evaporites. The enrichment of trace metals V and Cr, and the depletion of Co, Ni, and Mn suggest reducing diagenetic conditions in the initial sediment. Beluga calc-silicate rock is strongly depleted in REE (Total REE ~ 18 µg/g). Oxy-dravite δ11B (+3.9 ± 0.7‰) is consistent with a marine boron source. The oxygen isotope composition of corundum (δ18OVSMOW = 16.4 ± 0.1‰) is comparable to that of corundum in marble or desilicated pegmatite associated with marble. Phlogopite and muscovite 40Ar/39Ar ages and calculated closure temperatures 28  (considered estimates) are ca. 1640 Ma (Tc = 455 to 515 °C) and 1510 Ma (Tc = 410 to 425 °C), respectively. In the Lake Harbour Group, the most prospective areas for gem corundum exploration are expected to be contiguous to the thrust fault separating the Lake Harbour Group and Narsajuaq terranes, where the retrograde, amphibolite facies overprint of the granulite peak assemblages was most pervasive. 3.2 Chapter introduction Gem-quality blue corundum (sapphire) accounts for a significant portion of the gemstone market and prices continue to rise in response to demand (Shor & Weldon 2009, Genis 2016).  For this reason, there is considerable interest in understanding the genesis of gem-quality corundum and constraining the types of environments in which it forms. Improved exploration methodologies based on sapphire genetic models will aid in the development of a Canadian colored gemstone industry that will be competitive on the world market and could be applied in exploration efforts worldwide. Gem corundum deposits have been found in syenite, monzonite, kimberlite, lamprophyre, basalt (xenocrysts), gneiss, amphibolite, marble, skarn, and various contact metasomatic rocks (Simonet et al. 2008, Dzikowski et al. 2014, Giuliani et al. 2014).  Gem-quality blue and colorless corundum was discovered in 2002 near Kimmirut, southern Baffin Island, Nunavut, Canada, and subsequent exploration led to the discovery of blue, colorless, yellow, and pink gem corundum showings (True North Gems 2007). These occurrences are the first reported examples of gem corundum hosted in scapolite-rich calc-silicate pods in marble.  The present study examines two calc-silicate pods, the Beluga deposit, which is blue-corundum-bearing, and the Bowhead occurrence, a nearby calc-silicate pod in which corundum is rare, with the objective of determining the cause and timing of mineralization, and the nature of the protolith. 29  3.3 Exploration History The area around the Beluga and Bowhead calc-silicate pods was extensively explored by True North Gems, Inc. (TNG) from 2002 to 2009 after the initial discovery of the Beluga sapphire occurrence in 2002 by Seemeega and Nowdluk Aqpik, prospectors from Kimmirut.  Ultraviolet light prospecting for fluorescent scapolite assisted in the discovery of 8031 scapolite showings in and around the Beluga Project property, in addition to 45 named spinel outcrops, and 44 named corundum localities (some of which consist of multiple contiguous showings).  The distribution of corundum, scapolite, and spinel showings is plotted on a geological map of the TNG property in Fig. 3.1.   Several corundum occurrences produced notable gems as part of the exploration and deposit assessment work: the Beluga South occurrence produced 34 yellow sapphire gemstones totalling 6.98 carats, notably including two yellow stones weighing 1.47 carats and 1.09 carats (Fig. 3.2A), in addition to colorless and pale blue stones. The Aqpik occurrence produced two virtually flawless colorless stones, 2.50 carats and 2.59 carats in weight, and a pale blue, lightly included 7.81 carat gemstone (Fig. 3.2B).  Some of the Aqpik rough sapphire turned blue as a result of heat treatment.  One heat-treated, rich blue cushion-cut stone weighs 2.43 carats (Fig. 3.2A).  The Beluga occurrence is the most important gem corundum occurrence found in Nunavut to date, containing grades of gem sapphire rough between 33 g/t (4.29 t bulk sample) and 19 g/t (22.5 t bulk sample). Colorless, pale blue, and more commonly saturated to dark blue corundum occurs as crystals up to ca. 7 cm long.  One of the more notable stones in the bulk sample is a 1.17 carats, deep blue, extra fine sapphire gemstone (Figs. 3.2) showing even color when viewed through the table facet, but which in fact is primarily colorless with a central, dark blue patch (Wilson 2014).  Lastly, pinkish corundum was found at one location. 30  In total, 2607 polished corundum gemstones, totalling 169.95 carats (33.99 g), were cut from gem rough recovered by regional and bulk sampling.  The great majority of these gemstones originated from the Beluga deposit. 3.4 Results 3.4.1 Outcrop descriptions The Beluga (N 62.828734°, W 69.894072°; Fig. 3.3) and Bowhead (170 m to the SSW) occurrences are marble-hosted calc-silicate pods with surface exposures of 4.2 × 3.7 m and 2 × 1.5 m, respectively.  Contacts with the host marble are sharp, with the exception of parts of the Beluga pod, where euhedral crystals of dark brownish-purple diopside (1-3 cm) or yellowish-grey scapolite (1-5 cm) occur in very coarse pale-orange calcite between the calc-silicate rock and marble.  The rock in the calc-silicate pods is beige with abundant 1 to 6 cm spots of brown mottling.  Beige areas are primarily composed of scapolite, albite, muscovite, calcite, and corundum at Beluga, and scapolite-nepheline-calcite at Bowhead.  The brown mottling is phlogopite-rich with oligoclase and diopside.  The relative abundance of the light- and dark-coloured assemblages is variable: for example, at Beluga, the darker assemblage ranges between 10 and 90% on a decimeter scale but overall represents roughly half of the total rock volume based on visual estimation.  3.4.2 Petrography and mineral compositions 3.4.2.1 Beluga occurrence The calc-silicate rock at Beluga is coarse to very coarse-grained (5 mm to > 5 cm) and commonly consists of randomly oriented centimeter-scale crystals, occasionally with maximum 31  dimensions ca. 4 - 7 cm.  The calc-silicate rock can be divided into three distinct mineral assemblages (Fig. 3.4): (1) dark areas composed of phlogopite, oligoclase, diopside, and subordinate calcite; (2) light-colored, scapolite-rich zones, where scapolite is frequently altered to a mixture of fine-grained silicate minerals, generally separating the first assemblage from the corundum-bearing assemblage; and (3) pods or zones of light-colored corundum-albite-muscovite-calcite assemblage containing small amounts of graphite and pyrrhotite. Dark brown, Fe-, F-, and Ti-bearing phlogopite (Table 3.1), pale grey oligoclase (Ab80; Table 3.2), and subordinate calcite occur as graphic oriented intergrowths ranging in size from several millimeters to 2 cm across. Phlogopite-oligoclase symplectite up to 3 mm across is uncommon.  The oriented intergrowths form coronae around irregularly-shaped grains of purplish-brown Al-bearing, Si-poor diopside (Fig. 3.5) from several millimeters to 3 cm in size (0.31 Al pfu, 0.12 Na pfu; Table 3.3), and are uncommonly pseudomorphic after diopside.  The replacement of diopside by phlogopite-oligoclase is substantial (ca. 30-80%) with considerable local variation, and small relict diopside grains occur sparsely in the phlogopite-oligoclase.  Minor pale-orange calcite is present in the coronae and is infrequently visible in hand sample.  Titanite (<0.5 mm), zircon (Table 3.4), and apatite (0.1 mm) are uncommon.  The phlogopite-oligoclase intergrowths are locally associated with 1-3 cm masses of equigranular, medium-grained oligoclase with subordinate phlogopite and calcite.  Anhedral, coarse-grained to very coarse-grained crystals of pale yellowish-grey to grey scapolite occur around the phlogopite-rich coronae.  The scapolite averages 57 mol. % meionite component with considerable variation (Me50 – Me62; Table 3.5) and minor silvialite component (average 0.014 SO4 pfu).  Much of the scapolite is pervasively replaced by milky white, fine-32  grained mixtures (Fig. 3.6) of the following mineral phases, identified via X-ray powder diffraction: albite with variable quantities of prehnite, analcime, and thomsonite, and small amounts of calcite and muscovite.  The altered scapolite retains its cleavage in hand sample.  Uncommon, microscopic fractures cutting scapolite contain albite, calcite, and probable Mg-chlorite; other such fractures contain analcime.  Unaltered scapolite fluoresces bright yellow when exposed to long- and short-wave ultraviolet radiation.  Pristine scapolite, easily seen when exposed to UV, is generally restricted to the periphery of phlogopite-oligoclase mineralization and as isolated ‘pods’ within the light-coloured zones, where intensely altered scapolite is always contiguous to corundum-bearing zones (Figs. 3.6, 3.7).  Phlogopite crystals associated with scapolite, occurring on the periphery of phlogopite-oligoclase intergrowths, have well-developed basal faces and no embayed grain boundaries with scapolite, which suggests that these minerals form a stable assemblage. Small pods and zones, ranging in size from 0.5 × 1 cm to more than 4 × 10 cm, are composed of fine- to coarse-grained grey end-member albite (Ab98; Table 3.2), silvery-grey muscovite (Table 3.1), idiomorphic blue corundum, pale-yellow calcite (Fig. 3.6), subordinate graphite, and uncommon grains of anhedral pyrrhotite (<6 mm).  Oxy-dravite (Table 3.6) occurs very uncommonly at Beluga, and was only a significant constituent of a corundum-bearing pod in one sample, where it occurs as a 40 × 32 mm friable medium-brown mass near two dark brown, short-prismatic, euhedral crystals of the same mineral (up to 32 mm in length).  The corundum-bearing zones locally constitute up to 70% of the rock volume, estimated from visual inspection of outcrop.  Parts of the calc-silicate rock with high abundance of phlogopite-oligoclase-diopside (~90% of rock volume) tend to have little or no corundum mineralization.  The corundum-bearing zone contains open cavities where the rock surface, including euhedral 33  corundum, is entirely covered by a 1-2 mm thick coating of prismatic thomsonite with local analcime and grey, scalenohedral calcite.  Some cavities contain thin thomsonite prisms up to 8 mm in length with analcime and calcite crystals no larger than 2.5 mm.  LeCheminant et al. (2005) noted thomsonite seemingly penetrating a corundum crystal.  The surface of those corundum crystals enclosed in albite-muscovite-calcite and those enclosed in thomsonite are identical in morphology and show no corrosion or alteration. The corundum crystals are euhedral and tapered along the c axis; their faces are striated perpendicular to c.  Most crystals have dimensions in the range of 8 - 14 mm long and 1 – 4 mm wide, but larger crystals are not infrequent (2-4 cm long) and length:diameter aspect ratios range from approximately 3:1 to 9:1.  The largest crystal recovered intact at Beluga measured 7.7 × 2.1 cm (LeCheminant et al. 2005).  The corundum is royal blue, dark blue, or dark greyish blue in hand sample.  When observed in cut cross-sections or gemstones, the corundum varies from colorless and pale blue to bright blue, dark blue, or dark blue with a grey tinge in well-defined oscillatory zoning patterns and irregular sector zoning.  The corundum has good transparency and is sparsely included by calcite and apatite.  Gem quality is principally controlled by the degree of fracturing.  The corundum at Beluga contains an average of 0.08 wt. % TiO2 (range 0.00 – 0.30 wt. %) and 0.07 wt. % FeO (range 0.02 – 0.13 wt. %), and there is a weak positive correlation between Fe and Ti content (correlation coefficient = 0.26; see Dzikowski 2013 for spot analyses data). Rare rutile, sanbornite, thorianite, monazite, and uraninite are noted from the Beluga calc-silicate pod (Dzikowski 2013, LeCheminant et al. 2005). 34  3.4.2.2 Bowhead occurrence Mineralization in the Bowhead calc-silicate pod is relatively similar to that at Beluga with the notable exception that it is largely devoid of corundum mineralization whereas nepheline is common.  LeCheminant et al. (2005) noted trace corundum in the surface exposure.  The rock at Bowhead is composed of coarse to very coarse-grained phlogopite-oligoclase intergrowths (partly replacing diopside), diopside, scapolite, calcite, and nepheline.  The diopside and phlogopite at Bowhead are slightly poorer in Ti and Fe relative to Beluga (Tables 3.1 & 3.3).  The oligoclase is slightly more calcic than that at Beluga (Ab76; Table 3.2), and the same is true of scapolite (average Me60, range Me50-Me67; Table 3.5).  Both nepheline and scapolite have irregular curved boundaries with calcite.  Some scapolite grains appear to embay nepheline. Scapolite rims of variable thickness, some of which are very thin and difficult to observe in low magnification, occur around nepheline grains contiguous to calcite (Fig. 3.8).  Multiple nepheline grains are proximal to phlogopite-oligoclase intergrowths bordering diopside.  The extent of diopside replacement by phlogopite-oligoclase in the nepheline-bearing Bowhead samples (~15-30%) is inferior to the average level of replacement at Beluga and calcite is a major component of the nepheline-bearing rock, whereas it is less abundant at Beluga.  A single, 50 µm-wide nodule of muscovite (Table 3.1) and end-member albite (Ab98; Table 3.2) was found in scapolite. 3.4.3 Whole rock composition – Beluga deposit The Beluga calc-silicate pod is primarily composed of SiO2 (Table 3.7; ca. 46 wt. %), CaO (17 wt. %), Al2O3 (14 wt. %), and MgO (10 wt. %).  The rock is richer in Na than K (average Na2O = 2.71 wt. %, K2O = 2.02 wt. %, Na/K = 2.04) and contains 2.02 wt. % total Fe and 0.98 wt.% TiO2.  The atomic ratio of Al/Si is 0.36.  Loss on ignition is on the order of 5%, where CO2 is the principal volatile.  The rock is relatively enriched in V (Table 3.8; 237 µg/g) 35  and, to a lesser degree, Cr (108 µg/g).  The Beluga occurrence is very poor in Co (2.5 µg/g), Ni (5 µg/g), and Cu (5 µg/g).  Fluorine is relatively abundant (1260 µg/g).  The concentration of B is ~ 80 µg/g, and Cl ~545 µg/g.  The average total rare earth element (REE) concentration is relatively low, 18.1 µg/g, and the rock is slightly enriched in LREE with a smooth, flat HREE chondrite-normalized signature and variable negative Eu anomaly (see discussion). 3.4.4 Boron and oxygen isotope compositions The oxygen isotopic composition (δ18OVSMOW) of corundum from the Beluga occurrence is 16.4 ± 0.1‰.  Oxy-dravite has an average boron isotope composition (δ11B) of +3.9 ± 0.7‰ (Table 3.9). 3.4.5 Zircon U-Pb geochronology Zircon was recovered from the phlogopite-oligoclase portion of rock from the Beluga occurrence.  The zircon in the sample displays a range of morphologies, including coarse, irregular grains with rounded edges and sharp-faceted, stubby prismatic to equant grains.  All the grains are clear to translucent but range from medium to dark brown in color.  Five single grains (mostly ca. 130 µm in maximum dimension) were selected for analysis, including one coarse (>180 μm diameter), dark brown, translucent grain fragment with all rounded or broken edges (fraction A); two single, clear, stubby euhedral, prismatic brown grains (fractions B & C); and two single, clear, slightly elongate brown prismatic grains (fractions D & E).  The five zircon fractions contain relatively high U concentrations (608-1258 µg/g) and have low Th/U ratios (0.06-0.12; Table 3.10).  The analyses are slightly to strongly discordant (2.4-13.3%) but define a linear array, interpreted to reflect mainly recent Pb loss from a single age of metamorphic zircon, with calculated upper and lower intercepts (calculated using Model 1 of Ludwig 2003) of 1782.5 ± 3.7 Ma and -6 ± 140 Ma, respectively, with calculated MSWD = 0.23 (Fig. 3.9). 36  3.4.6 Ar-Ar ages of mica and estimate of Tc Two 40Ar-39Ar data collections from Beluga phlogopite (Appendix A) produced: (1) a flat spectrum yielding a plateau age of 1635.9 ± 8.4 Ma (Fig. 3.10A) where data corresponding to 91.4% of the 39Ar volume was used, and (2) a non-ideal saddle-shaped spectrum with a flat minimum yielding a plateau age of 1646.8 ± 8.6 Ma (Fig. 3.10B) where 63.7% of the data was used.  Two analyses of muscovite (Appendix A) yielded the following plateau ages: (1) 1511.7 ± 8.4 Ma using 34% of the 39Ar volume (Fig. 3.10C), and (2) 1510.4 ± 8.3 Ma using 78% of the 39Ar volume (Fig. 3.10D). Closure temperatures were calculated using CLOSURE v1.2 (Brandon et al. 1998) and different sets of activation energies and frequency factors for phlogopite and muscovite (Table 3.11).  Using the activation energies and frequency factors of Villa (2010) and Harrison et al. (2009), we estimated the cooling rate to be on the order of ca. 0.35-0.75 °C/Ma, and the closure temperatures of phlogopite and muscovite to be ca. 455-515 °C and 410-425 °C, respectively. The accuracy of these results is limited by the quality of the models, the validity of underlying assumptions (size of diffusion domain), and the use of only two minerals as geothermometers. 3.5 Discussion 3.5.1 Paragenetic sequence and metamorphic history Butler (2007) suggested the following possible nepheline-forming reactions for rocks on Aliguq island: dolomite + albite  diopside + nepheline + 2 CO2 (1) phlogopite + 3 calcite + 3 albite  3 diopside + 3 nepheline + K-feldspar + 3 CO2 + H2O (2) Based on the inclusion of nepheline by diopside, phlogopite, and forsterite, Butler (2007) ruled-out reaction (1), since it occurs at lower metamorphic grade than a forsterite-bearing 37  assemblage.  The implied late formation of nepheline and its intergrowth with K-feldspar led Butler (2007) to infer that nepheline formed from reaction (2).  A thermodynamic model of the reaction presented by Butler (2007) suggests that at 8 kbar pressure, reaction 2 only occurs at XCO2 < 0.15.  The reaction produces CO2 at a ratio of 3:1 relative to H2O, and therefore the reaction would be self-limiting under these P-T conditions unless XCO2 was buffered by a large supply of H2O-dominant metamorphic fluid.  In the Beluga and Bowhead calc-silicate pods, the break-down of diopside to a mixture of phlogopite, oligoclase, and calcite clearly implies the retrograde reversal of the peak metamorphic assemblage in reaction 2. Thermodynamic modelling of reaction 2 in a TWQ v1 P-T phase diagram (Fig. 3.11A; Berman 1988, 1991) places the phase boundary out of range of the peak metamorphic grades, thus low XCO2 (as per Butler 2007, Fig. 3.11B) is required to form the nepheline-bearing assemblage. Reaction 2 is significantly more consistent with the whole rock composition and mineral parageneses compared to reaction 1.  The expected whole rock composition based on the mineral proportions in reaction 1 differs significantly from the measured composition of Beluga calc-silicate rock: the abundance of K relative to Na (whole rock K/Na = 0.49) cannot be accounted for by the relatively small K concentrations found in nepheline (Bowhead nepheline K/Na = 0.16).  Potassium feldspar was not observed in samples from Beluga or Bowhead, but it has been noted in other corundum-bearing calc-silicate pods in the area (Hansen 2008). Therefore, the presence of K-feldspar in the peak metamorphic assemblage is inferred from its suggested role in metamorphic reactions, its coeval relationship with nepheline in calc-silicate rocks elsewhere in the LHG (Butler 2007), and since it can account for whole rock K concentrations higher than expected for the peak metamorphic paragenesis partly preserved in Beluga and Bowhead calc-silicate rock. 38  Zircon recovered from phlogopite-oligoclase coronae have a concordia intercept age of 1782.5 ± 3.7 Ma, interpreted as the age of zircon neocrystallization and break-down of the peak metamorphic assemblage during the post-D2 fluid incursion as suggested by St-Onge et al. (2007) in P-T conditions slightly lower than M2 metamorphism (720 °C, 6.2 kbar).  The timing of this retrograde mineralization is consistent with slow cooling rates estimated from phlogopite-muscovite ages (0.35-0.75 °C/Ma, estimated from running multiple iterations for closure temperatures in CLOSURE v1.2 calculations using phlogopite and muscovite). The extent to which the retrograde mineral assemblage replaces the peak assemblage (i.e., reversed reaction [2]) is dependent on the availability of CO2 and H2O in fluids introduced during this episode of retrograde metamorphism.  Scapolite appears stable with phlogopite since the phlogopite crystals possess well-developed basal faces, and neither mineral embays the other: thus we infer that the scapolite is part of the phlogopite-oligoclase-calcite assemblage, although since it is not present within the phlogopite coronae, it may have formed slightly later from destabilization of nepheline not consumed in the phlogopite-forming reaction.  The higher XCa of scapolite (Me50-67) relative to oligoclase (An19-24) is consistent with compositional data for coexisting scapolite-plagioclase pairs in metamorphic rocks, but F and Cl contents of phlogopite (XMg ≈ 0.93, F ≈ 1.39 – 1.96 wt. %, Cl = 0.02 wt. %) are significantly different from values in biotite associated with scapolite in these metasediments (XMg ≈ 0.6, F ≈ 0.6 wt.%, Cl ≈ 0.2 wt.%; Mora & Valley 1989). At the Bowhead occurrence, rims of scapolite separate nepheline and calcite; this suggests that during the first episode of retrograde metamorphism, scapolite formed from the reaction of nepheline with calcite.  The availability of NaCl in fluid stabilizes scapolite relative to plagioclase in the presence of calcite (Ellis 1978), but the effect of fluid Cl content on the relative stability of 39  nepheline-scapolite-calcite is unknown.  In corundum-bearing calc-silicate rock at Pitcairn, New York, rims of NaCl-bearing scapolite separating nepheline from calcite or apatite imply the formation of scapolite from the reaction of nepheline with calcite or apatite and CO2- and Cl-bearing fluid (P.M. Belley, unpublished data).  We suggest that scapolite formation at Beluga and Bowhead may have been limited by Cl availability, whereas fluid Na content may not be a controlling factor since the nepheline is Na-rich.  Therefore, the extent of scapolite formation at the expense of nepheline and calcite may be controlled by the availability of Cl introduced to the system concomitant with CO2- and H2O-influx during retrograde metamorphism. Pods and zones of albite, muscovite, calcite, and corundum are always surrounded by pervasively albitized (Ab98) scapolite.  A similar phenomenon is observed on a centimetre scale around small corundum-bearing zones in scapolite at Pitcairn, New York, but muscovite is not present at this locality (P.M. Belley, unpublished data).  This consistent spatial relationship, together with the presence of calcite in the fine-grained alteration, suggests that the carbonate-bearing scapolite is a reactant in corundum-forming reaction (3), below. Reaction (3) excludes graphite and pyrrhotite, two minerals occurring in the corundum-bearing assemblage at Beluga which probably formed as a result of carbonate and sulphate (released from scapolite break-down) reduction, respectively.  Scapolite (Me57) + 0.84 nepheline + 0.60 H2O + 2.02 CO2  0.76 corundum + 2.35 calcite + 2.14 albite + 0.60 muscovite + 0.66 Cl (3)    40  Reaction (4), an approximation of reaction (3), was used to model in TWQ due to limitations in scapolite solid solution and the inclusion of muscovite in the reaction:  meionite + 3 nepheline + 3 CO2  3 corundum + 4 calcite + 3 albite (4)  The phase boundary for reaction (4) is in the 650-700 °C range for 2.5-4 kbar pressure (Fig. 3.11A).  The phase boundary temperature for the break-down reaction of scapolite and nepheline with CO2 is probably overestimated by the TWQ thermodyamic model, since Kimmirut scapolite is richer in Na (Me50-67) and end-member meionite becomes stable at higher temperature relative to Na-bearing scapolite (Goldsmith & Newton 1977).  Moreover, the TWQ modeled reaction does not take into account the formation of muscovite, an important constituent of the albite-calcite-corundum assemblage at Beluga, however this may not be an important control on mineralization, since muscovite is absent in a similar retrograde corundum-calcite-albite assemblage at Pitcairn, NY. Zeolites, whose presence is characteristic of low-grade metamorphism, formed in corundum-bearing zones at Beluga relatively late (<1500 Ma, since mica closure temperatures indicate that temperatures were above zeolite facies conditions at this time), and are accompanied by the dissolution of calcite, or the break-down of scapolite or possibly relict nepheline, resulting in the creation of open space.  Uncommon oxy-dravite crystals occur near the border of phlogopite-oligoclase and the corundum-bearing zone. Due to the wide P-T stability range of tourmaline (van Hinsberg et al. 2011) and the lack of mineral textures that could provide information on its paragenesis, the position of oxy-dravite in the paragenetic sequence could not be determined. 41  3.5.2 Controls on corundum genesis At Kimmirut, corundum was formed  by the following sequence of events, determined from the petrography and modeled reactions discussed in section 3.5.1: (1) formation of a nepheline-diopside-rich peak metamorphic assemblage; (2) high-temperature partial retrogression of this assemblage where some nepheline is preserved and where scapolite forms at its expense; and (3) introduction of CO2-H2O-bearing fluid at slightly lower temperature, causing nepheline and scapolite to react, forming albite, calcite, corundum, and muscovite.  Muscovite formation is expected to be controlled by the availability of K (from nepheline) and H2O (fluid) and is probably not essential to the corundum-forming reaction.  Corundum mineralization is extensive at the Beluga occurrence whereas it is rare at Bowhead, where the nepheline is largely unaltered.  Since the occurrences are separated by a distance of only 170 m, it is evident that local variations play an important role in controlling the reaction. Corundum forms from the break-down of nepheline + scapolite is predicted to occur in a narrow temperature range with upper and lower limits controlled by nepheline + scapolite stability and corundum vs. Al-silicate stability, respectively.  Al-silicate forms by reaction (5):  2 meionite + 3 nepheline + 6 CO2  8 calcite + 6 Al2SiO5 + 3 albite (5)  A TWQ thermodynamic model (Fig. 3.11A) indicates a 70-90 °C window in which nepheline and meionite would break down to form corundum rather than Al-silicate for expected pressures in a barrovian retrograde P-T path.  The actual position of this window in P-T space could only be determined with improved thermodynamic models for Na-bearing scapolite. 42  3.5.3 A possible magmatic origin? LeCheminant et al. (2005) proposed that the Beluga and Bowhead calc-silicate pods formed from syenitic magma that intruded, and were contaminated by marble during D2 deformation.  Partly contaminated syenitic intrusions have not been noted in the area, but they would be expected given the widespread occurrence of scapolite-bearing and corundum-bearing calc-silicate pods in LHG marble near Kimmirut. Some scapolite-rich pods are small (ca. 1 m in maximum dimension) and isolated within the regional marble (Hansen 2008).  Moreover, the contacts between marble and the calc-silicate rocks are generally sharp with no apparent zoning.  The author (PMB) has observed multiple granite pegmatites in marble throughout the Central Metasedimentary Belt, Grenville Province, where metamorphic grades were in upper amphibolite to granulite facies, and these dikes typically have no to minor contamination in the form of Ca-Mg-bearing silicate minerals.  The dikes typically have thin metasomatic aureoles at their contact with marble (e.g., tremolite around zircon-rich pegmatite near Bryson, Québec), or are contiguous with coarsely crystalline vein- or pod-like bodies of clinopyroxene and feldspar/scapolite with calcite-rich cores (e.g., rocks in the Lac Tortue area, ZEC Bras-Coupé Désert, Québec).  Compared to these examples, the Beluga and Bowhead calc-silicate pods are remarkably uniform, despite local variations in mineral abundance.  Th/U ratios in Beluga zircons (Th/Uavg = 0.08) are within the range considered to be characteristic for metamorphic zircon (Th/U < 0.1) by Rubatto et al. (1999), although Möller et al. (2003) cautioned against attributing the origin of zircon using Th/U.  Lastly, the paragenetic sequence of the Beluga and Bowhead calc-silicate pods appears most consistent with a metamorphic origin where the peak metamorphic assemblage was subjected to two distinct stages of high-temperature retrograde 43  metamorphism, and the inferred peak P-T formation of nepheline was noted in carbonate-bearing metasediments in another region of the LHG (Butler 2007). 3.5.4 Nature of the protolith Beluga calc-silicate rock plots between LHG metapelites and psammites in Al/Si and Na/K ratios (Fig. 3.12A), is slightly Fe-depleted relative to shale and metapelites (Fig. 3.12B), and contains a TiO2 concentration (~1 wt. %) consistent with a shale/pelitic component (Fig. 3.12D).  The rock is significantly richer in CaO and MgO than LHG clastic metasediments and intrusive rocks but with similar concentrations to LHG lapis lazuli metaevaporite layers in marble (Fig. 3.12C).  These data suggest a protolith of mixed clastic and carbonate composition, where the clastic component is an intermediate between the pelitic and psammitic sediments, i.e., a silty clay. While a small quantity of halite could explain the greater abundance of Na relative to K, the relative abundance of these elements at Beluga is consistent with an intermediate between LHG metapelite and psammite, and the Na/K value at Beluga (~2) is significantly lower than that at the Soper River lapis lazuli metaevaporite (~ 5 to 19; Hogarth & Griffin 1978). Metamorphic reactions resulted in significant decarbonation of the rock, but initial carbonate content can be estimated by subtracting an estimated MgO and CaO siliciclastic contribution (an intermediate of LHG metapelite-psammites, 2 wt. % CaO, 3-6 wt. % MgO) from the total bulk CaO and MgO, and recalculating the difference as calcite and dolomite with a second iteration correcting assumed initial CaO and MgO values for the amount of original allochthonous material (i.e., 1.6 wt. % CaO, 2.6-5.2 wt. % MgO).  The resulting estimation is a protolith with ca. 19 wt. % dolomite and 13 wt. % calcite to 29 wt. % dolomite and 7 wt. % calcite.  Therefore, the most likely protolith based on major element bulk composition is a silty or sandy dolomitic argillaceous marl, although a thinly interlayered dolostone-shale sequence is equally possible 44  given the coarse size of metamorphic recrystallization. It should be noted that the initial relative abundance of Na vs. K may have been affected by diagenetic processes, notably albitization of detrital feldspars (e.g., Baccar et al. 1993).  The trace-element composition of Beluga calc-silicate rock is not as simple to interpret and it is important to consider that the complex metamorphic history of the deposit may have obscured the original trace element signature in the protolith.  Vanadium (237 µg/g) is significantly enriched relative to Cr (108 µg/g), while Co, Ni, and Cu (≤ 5 µg/g) are extremely depleted. The V concentration at Beluga is similar to V-rich LHG metapelite and psammite, while Cr is consistent with shale, argillite, or marl, and Co, Ni, and Cu concentrations are closer to that of marble (Fig. 3.13).  According to Tuttle et al. (2000), “aluminum and titanium are mostly bound to phases that are relatively unreactive in marine environments; therefore, both of these elements provide a good estimate of the amount of allochthonous detritus.”  Using this assumption, V, Cr, and Ni concentrations in Beluga rock, LHG metapelite, and psammite are plotted against the expected clastic contribution of trace elements in average shale (Fig. 3.14).  The comparison with shale suggests significant V enrichment, typical detrital Cr contribution, and significant Ni (and by proxy, Co and Cu) depletion.  One LHG psammite (95-C084 of Thériault et al. 2001) shows similar V-enrichment, expected Cr, and a weaker Ni-depletion (29 µg/g). A study of transition metal behavior in response to different early diagenetic environments of modern sediments (Shaw et al. 1990) demonstrated the trapping of Ni and Co with manganese oxides, which are enriched and preserved in the oxic zone of sediments and released under reducing conditions.  In contrast, near-anoxic reducing conditions (e.g., in organic-matter-bearing sediment) favor the enrichment and preservation of Cr, V, and Mo, where 45  accumulation of Cu is moderately enriched in response to reducing conditions and closely correlated with biogenic material flux (i.e., the Cu binding capacity of sediment decreases in slowly accumulating sediments where a smaller fraction of biogenic detritus survives). In strongly reducing H2S-rich environments which favour high V/(Ni+V) ratios, maximum molar V/(Ni+V) are on the order of ~ 0.85 (black shale-carbonate sequences of Wenger & Baker 1986) – significantly lower than that at Beluga (V/(Ni+V) ≈ 0.98). Vanadium is strongly enriched in the Beluga calc-silicate rock, and Cr well-preserved.  Molybdenum concentration is below detection level (<2 µg/g), but considering the low concentration in average shale (2 µg/g; Carmichael 1989) and significant dilution by the carbonate component in the protolith, Mo enrichment may not be detectable.  Low MnO (0.01 wt. %) relative to a sample of LHG marble (0.17 wt. %; Butler 2007), average shale, and average marine carbonate (0.08 wt. %, 0.07 wt. %; Carmichael 1989) is consistent with the proposed mechanism of Co and Ni depletion, but many LHG metapelites and psammites contain low MnO with highly variable Ni (Fig. 3.13). The extremely low REE concentrations relative to other lithologies in the LHG and Meta Incognita peninsula (Fig. 3.15) could partly be explained by early, reducing diagenetic conditions since REE also concentrate on Mn oxides, although REE could be preserved in phosphate in anoxic conditions (Takahashi et al. 2015).  The V-enriched, Ni-depleted LHG psammite is far richer in total REE (~335 µg/g) relative to Beluga (~18 µg/g), which does not support this hypothesis.  In LHG metapelites, concentrations of V, Ni, and Cr show some degree of positive correlation.  Diagenetic processes could potentially dissolve phosphate minerals and mobilize REE.  Given the complex metamorphic history of the Beluga calc-silicate rock, the attribution of early diagenetic processes to observed trace element profiles carries high uncertainty. 46  The oxygen isotope composition of corundum at Beluga is slightly higher than for corundum in desilicated pegmatite in marble, and at the low end of the marble range (Fig. 3.16). A 18O value of +16.4‰ is consistent with the equilibration of oxygen isotopic composition with marble, and it is higher than for corundum in skarn developed in marble (Giuliani et al. 2014). 3.5.5 On evaporites Dzikowski (2013) suggested that the Beluga and Bowhead protoliths contained an evaporitic component.  Here, we compare the bulk composition of Beluga calc-silicate rock with the geochemistry of evaporite-bearing argillites, dolomitic marls, and mixed sulfate rocks.  These evaporite-bearing or evaporite-related sediments, in comparison with common platform sediments, have high Mg, high Mg/Ca ratios (except Ca-sulfate-rich rock), low Fe contents, high K with low Na (except halite-bearing rocks), and specifically in argillites, high Li, F, and B/Al (Moine et al. 1981). The Al-Mg-Ca geochemical signature of Beluga calc-silicate rock is within the range of platform marl and clay-shale, and just outside the compositional domain for evaporites and metaevaporites (Fig. 3.17). Hypothetically, the evaporite-associated sediment best matching the bulk major element composition of Beluga rock is argillaceous marl (clay/silt with dolomite and calcite) containing halite (Na > K), but Na-K contents are also consistent with a mixed clastic component (i.e., intermediate between LHG metapelites and psammites, Fig. 3.12A).  The boron concentration at Beluga (80 µg/g) is similar to that of average shale (100 µg/g; Carmichael 1989) but far lower than in argillaceous evaporitic rocks (200-400 µg/g; Moine et al. 1981).  The whole-rock B measurements at Beluga may be affected by localized oxy-dravite mineralization (i.e., oxy-dravite is very uncommon, but one sample contains a ~8 cm cluster).  The concentration of Li at Beluga (15 µg/g) is far lower than the range for most evaporitic rocks (45 – 300 µg/g), however Li concentrations are low in Mg-clay-poor, dolomite-47  rich sedimentary rocks associated with evaporites (Moine et al. 1981).  Beluga F content (1260 µg/g), mostly held in phlogopite, is comparable to evaporitic argillite (1000-2000 µg/g; Moine et al. 1981) and approximately double that of shale (500-740 µg/g; Carmichael 1989).  Beluga calc-silicate rock is also enriched in Cl (~500 µg/g) relative to shale (160-180 µg/g; Carmichael 1989).  Fluorine-rich phlogopite and Cl-rich scapolite are retrograde assemblages: F and Cl are not highly compatible constituents within the major mineral phases in the peak assemblage (nepheline, diopside, K-feldspar), therefore it can be surmised that these elements were probably introduced with metamorphic fluids during retrograde metamorphism. The boron isotope composition of oxy-dravite is similar to tourmaline with B sourced from marine boron.  Due to the overlap of tourmaline δ11B in marble, evaporite-associated metapelite, and B-rich metaevaporite (Fig. 3.18), an evaporitic protolith cannot be inferred from the data. In summary, bulk rock major and trace element data are consistent with non-evaporitic marine platform sediments. There is no evidence to support an evaporitic origin of the protolith, although the possibility that the protolith contained a small evaporite component cannot be ruled out. 3.5.6 Implications for gem corundum exploration Since corundum mineralization is dependent on two episodes of amphibolite facies retrograde metamorphism in Beluga- and Bowhead-like rocks, parts of the LHG with pervasive amphibolite-facies retrograde overprinting of peak metamorphic assemblages would be more prospective than areas with well-preserved, peak granulite facies assemblages.  In the LHG, domains with pervasive amphibolite facies retrograde metamorphism are restricted to the 48  periphery of the thrust fault between the LHG and the Narsajuaq terrane (see Fig. 5 of St-Onge et al. 2000). Kimmirut-type gem corundum deposits are produced by a very specific P-T history and protolith composition, and therefore are probably rare compared to gem corundum deposits with simpler paragenetic sequences and more common protoliths (especially marble-hosted deposits, see Giuliani et al. 2014).  A corundum occurrence similar to the Kimmirut deposits is located in Pitcairn, New York, USA, where purplish-red, opaque corundum occurs with albite and calcite formed at the expense of scapolite and nepheline (unpublished data, P.M. Belley; Chamberlain et al. 2015). 3.6 Conclusions In Lake Harbour Group calc-silicate rocks near Kimmirut, Nunavut, corundum formation was made possible by three equally important sequential metamorphic reactions: (1) formation of nepheline, diopside, and K-feldspar (inferred) at peak metamorphic (granulite facies) conditions from a metamorphosed dolomitic argillaceous marl precursor; (2) partial retrograde replacement of the peak assemblage by phlogopite, oligoclase, calcite, and scapolite as a result of CO2-, H2O-, Cl-, F-bearing fluid influx at 1782.5 ± 3.7 Ma, ~30 Ma younger than the end of M2 metamorphism (therefore P-T < 720°C, 6.2 kbar); and (3) retrograde break-down of scapolite + nepheline (+ CO2 + H2O) to form albite, muscovite, corundum, and calcite.  As evidenced by the abundance of corundum at Beluga and its near absence at Bowhead (170 m away), conditions favouring the alteration of scapolite-nepheline were locally heterogeneous.  The corundum-forming reaction only occurs in a <100°C window with a lower limit determined by the formation of Al-silicate rather than corundum.  Thermodynamic models of these reactions are inconsistent with known metamorphic conditions in the LHG, but do not take into account 49  variations in the composition of major phases.  They are interpreted to be overestimates of phase boundary P-T since they exceed estimates based on the paragenetic sequence and the regional metamorphic history.  The inferred mineral reactions are in strong agreement with petrologic observations, mineral assemblages, and bulk composition. Phlogopite and muscovite 40Ar/39Ar ages and calculated closure temperatures are ca. 1640 Ma (Tc = 455 to 515°C) and 1510 Ma (Tc = 410 to 425°C), respectively.  The late formation of thomsonite, analcime, and prehnite at Beluga is expected to be considerably younger than the muscovite closure age. Local-scale and outcrop-scale field observations, combined with bulk and trace element geochemistry, suggests that the Beluga and Bowhead calc-silicate pods are metasedimentary.  The relative abundances of Ca, Mg, Al, Si, Ti, and Fe are consistent with dolomitic argillaceous marl, where the siliciclastic component is an intermediate between that of LHG metapelites and psammites (i.e., sandy or silty clay).  Significant V enrichment, high Cr, and very low Ni, Co, and Cu may be related to early diagenetic processes such as a slow sedimentation rate and reducing conditions caused by the presence of organic matter. The Li and B concentrations do not suggest an evaporite-related protolith, although concentrations of these elements in evaporite-associated sediments can be low.  The Na/K ratios are intermediate between LHG metapelites and psammites.  High F and Cl concentrations (ca. 1260 and 500 µg/g, respectively) appear to have been introduced by metamorphic fluid during retrograde metamorphism.  Tourmaline boron isotopes are consistent with a marine boron source but cannot be used to distinguish between normal marine and hypersaline environments.  Therefore, data do not suggest that the Beluga protolith was evaporite-related; i.e., Beluga whole rock geochemistry is within the range expected in non-evaporitic marine shelf sediments. 50  However, the possibility that an initial, minor evaporitic component was present in the protolith cannot be rejected.  We infer a similar origin for the Bowhead calc-silicate rock due to its mineralogical and geological similarity to the Beluga occurrence. In the Lake Harbour Group, the most prospective areas for gem corundum exploration are expected to be contiguous to the thrust fault separating the LHG and Narsajuaq terranes, where the retrograde, amphibolite facies overprint of the granulite peak assemblages is most pervasive.   51  3.7 Tables  Table 3.1: Average composition of phlogopite and muscovite from the Beluga and Bowhead calc-silicate pods. After Dzikowski (2013).   Wt.% phlogopite Beluga n = 41 phlogopite Bowhead n = 18 muscovite Beluga n = 12 muscovite Bowhead n = 2 SiO2 39.79 40.99 44.96 45.00 TiO2 2.53 1.84 0.04 0.04 Al2O3 15.58 15.01 37.39 38.47 Cr2O3 0.03 0.01 0.01 0.00 MgO 23.32 25.12 0.02 0.08 CaO 0.02 0.01 0.01 0.14 BaO 0.04 0.06 0.01 0.09 MnO 0.03 0.02 0.02 0.00 FeO 2.92 2.14 0.04 0.03 Na2O 0.27 0.31 0.81 1.76 K2O 10.56 10.63 10.68 8.96 F 1.39 1.96 0.00 0.00 Cl 0.02 0.02 0.02 0.04 H2O* 3.57 3.36 4.46 4.51 O=F,Cl -0.59 -0.83 0 -0.01 Total 99.48 100.65 98.46 99.11 Normalized to 12 anions with OH* + F + Cl = 2 Si (apfu) 2.820 2.860 3.020 2.985 Ti 0.135 0.097 0.002 0.002 Al (IV) 1.180 1.140 0.980 1.015 Al (VI) 0.121 0.094 1.980 1.993 Cr 0.002 0.001 0.001 -- Mg 2.463 2.613 0.002 0.008 Ca 0.002 0.001 0.001 0.010 Ba 0.001 0.002 0.000 0.002 Mn 0.002 0.001 0.001 -- Fe 0.173 0.125 0.002 0.002 Na 0.037 0.042 0.105 0.226 K 0.955 0.946 0.915 0.758 F 0.312 0.432 -- -- Cl 0.002 0.002 0.002 0.004 OH* 1.686 1.565 1.998 1.996 ∑cations 7.891 7.922 7.009 7.001  *Calculated, OH = 2 – F – Cl     52  Table 3.2: Average composition of plagioclase and nepheline from the Beluga and Bowhead calc-silicate pods. After Dzikowski (2013).    Wt.% oligoclase Beluga n = 42 albite Beluga n = 9 oligoclase Bowhead n = 8 albite Bowhead n = 5 nepheline Bowhead n = 20 SiO2 63.27 66.31 62.65 68.16 43.49 Al2O3 23.27 20.61 23.74 19.96 34.45 MgO 0.13 0.07 0.00 0.00 0.00 CaO 3.93 0.41 4.96 0.38 2.37 MnO 0.01 0.00 0.01 0.02 0.00 FeO 0.05 0.02 0.03 0.01 0.00 Na2O 9.23 11.61 8.82 11.33 15.69 K2O 0.19 0.11 0.07 0.10 3.92 Total 100.08 99.14 100.28 99.96 99.92 Norm. 8 O 8 O 8 O 8 O 4 O Si (apfu) 2.793 2.932 2.766 2.978 1.035 Al 1.211 1.074 1.235 1.028 0.966 Mg 0.009 0.005 -- -- -- Ca 0.186 0.019 0.235 0.018 0.061 Mn 0.000 -- 0.000 0.001 -- Fe 0.002 0.001 0.001 0.000 -- Na 0.790 0.995 0.755 0.960 0.724 K 0.011 0.006 0.004 0.006 0.119 Or 0.01 0.01 0.00 0.01  Ab 0.80 0.98 0.76 0.98  An 0.19 0.02 0.24 0.02     53  Table 3.3: Average composition of diopside from the Beluga and Bowhead calc-silicate pods. After Dzikowski (2013).   Beluga Bowhead   Beluga Bowhead Wt.% n = 18 n = 9  Normalized to 6 anions SiO2 51.24 51.99  Si (apfu) 1.856 1.888 TiO2 1.29 0.78  Ti 0.035 0.021 Al2O3 7.25 5.76  Al 0.310 0.247 Cr2O3 0.02 0.01  Cr 0.001 0.000 MgO 14.24 15.01  Mg 0.769 0.813 CaO 22.52 23.28  Ca 0.874 0.906 MnO 0.04 0.03  Mn 0.001 0.001 FeO 1.59 1.20  Fe 0.048 0.036 Na2O 1.69 1.56  Na 0.119 0.110 Total 99.88 99.62  ∑cations 4.013 4.022      Table 3.4: Composition of zircon from the phlogopite-oligoclase portion of Beluga calc-silicate rock.  Wt.% n = 2 SiO2 33.54 ZrO2 65.34 HfO2 1.36 La2O3 0.01 Ce2O3 0.01 Pr2O3 0.07 Nd2O3 0.02 Sm2O3 0.13 Gd2O3 0.03 ThO2 0.05 UO2 0.00 Total 100.56      54  Table 3.5: Average composition of scapolite from Beluga and Bowhead calc-silicate pods with standard deviation and minimum/maximum me% compositions. See Dzikowski (2013) for full data set.   Beluga σ Min.** Max.**  Bowhead σ Min.** Max.** Wt.% n = 67     n = 15 SiO2 49.04 1.78 49.90 46.84  47.95 1.35 49.54 45.44 Al2O3 26.30 1.00 25.52 27.09  26.21 0.75 25.31 27.58 MgO 0.06 0.22 0.00 0.00  0.00 0.01 0.00 0.00 CaO 14.34 1.77 13.03 15.13  15.10 1.17 14.02 17.38 MnO 0.01 0.02 0.00 0.00  0.01 0.02 0.00 0.01 FeO 0.02 0.03 0.10 0.00  0.02 0.02 0.00 0.08 Na2O 5.52 0.93 6.12 4.97  5.31 0.66 5.87 4.16 K2O 0.45 0.18 0.49 0.39  0.33 0.08 0.39 0.24 SO3 0.12 0.05 0.08 0.09  0.12 0.08 0.20 0.06 Cl 2.59 0.92 3.65 2.30  2.65 0.57 3.24 1.66 CO2* 1.82 1.02 0.31 1.90  1.71 0.70 0.71 2.66 O=Cl -0.58 0.21 -0.82 -0.52  -0.6 0.13 -0.73 -0.37 Total 99.69 0.81 98.37 98.19  98.81 0.49 98.55 98.9 Normalized to Al + Si = 12 Si (apfu) 7.353  7.487 7.136  7.189  7.490 6.996 Al 4.647  4.513 4.864  4.631  4.510 5.004 Mg 0.013  -- --  0.000  -- -- Ca 2.304  2.095 2.470  2.426  2.271 2.867 Mn 0.001  -- --  0.001  -- 0.001 Fe 0.003  0.013 --  0.003  -- 0.010 Na 1.605  1.780 1.468  1.544  1.721 1.242 K 0.104  0.094 0.076  0.076  0.075 0.047 SO4 0.014  0.009 0.010  0.014  0.023 0.007 Cl 0.658  0.928 0.594  0.673  0.830 0.433 CO3* 0.328  0.063 0.396  0.313  0.147 0.560 Me(%) 57  50 62  60  50 67 *Calculated assuming CO3 + Cl + SO4 = 1 ** Examples of minimum and maximum Me(%) analyses   55  Table 3.6: Average composition of oxy-dravite at the Beluga occurrence.   Wt.% Beluga n = 7 Site Normalized to 15 Y + Z + T cations SiO2 36.18 T Si (apfu) 5.844 TiO2 0.36 T Al 0.156 B2O3a 10.76  Ba 3.000 Al2O3 37.10 Z Al (Z) 6.000 V2O3 0.03 Y Ti 0.044 Cr2O3 0.01 Y Al 0.906 FeO 1.14 Y V 0.004 MnO 0.00 Y Cr 0.001 ZnO 0.02 Y Fe2+ 0.154 MgO 7.84 Y Mn -- CaO 0.46 Y Zn 0.002 Na2O 2.82 Y Mg 1.888 K2O 0.04 X Ca 0.080 F 0.30 X Na 0.883 Cl 0.00 X K 0.008 O=F,Cl -0.13 X Vacancy 0.029 Total 99.68 V OHb 2.950   V Ob 0.050   W F 0.153   W Cl 0.000   W Ob 0.847 aBoron calculated assuming ideal 3 B apfu bOH-O calculated by cation charge balance   56  Table 3.7: Whole rock major element composition, Beluga occurrence. After Dzikowski (2013).  Wt.% BA-1 BA-2 BA-3 BA-4 BA-5 Average SiO2 45.19 45.86 46.07 45.37 46.54 45.81 TiO2 0.92 1.07 1.03 0.93 0.95 0.98 Al2O3 15.26 11.47 13.3 15.32 14.92 14.05 Cr2O3 0.01 0.01 0.01 0.01 0.01 0.01 Fe2O3 1.47 1.83 1.53 1.43 1.47 1.55 FeO 1.16 1.42 1.16 1.16 1.09 1.2 MnO 0.02 0.03 0.03 0.02 0.02 0.02 MgO 9.28 11.46 10.53 9.59 9.87 10.15 CaO 15.82 18.86 17.62 16.07 16.92 17.06 SrO 0.02 < 0.01 0.01 0.03 0.02 0.02 BaO 0.02 0.01 0.02 0.02 0.02 0.02 Na2O 2.96 2.43 2.68 2.94 2.55 2.71 K2O 2.45 1.55 1.86 2.37 1.88 2.02 P2O5 0.03 0.02 0.02 0.02 0.02 0.02 LOI 5.58 4.56 4.37 5.1 3.95 4.71 Total 100.19 100.58 100.24 100.38 100.23 100.33 H2O- 0.04 0.01 0.06 0.04 0.02 0.03 H2O+ 0.89 0.66 0.66 1.12 1.22 0.91 C 1.15 0.95 0.93 1.05 0.66 0.95 CO2 4.2 3.5 3.4 3.8 2.4 3.46 Mol.%       P 0.009 0.006 0.006 0.006 0.006 0.006 Si 15.338 15.805 15.832 15.293 15.900 15.634 Ti 0.235 0.277 0.266 0.236 0.244 0.252 Al 6.104 4.659 5.387 6.086 6.008 5.651 Cr 0.003 0.003 0.003 0.003 0.003 0.003 Fe3+ 0.375 0.475 0.396 0.363 0.378 0.398 Fe2+ 0.329 0.409 0.333 0.327 0.311 0.343 Mn 0.006 0.009 0.009 0.006 0.006 0.006 Mg 4.696 5.888 5.395 4.819 5.027 5.164 Ca 5.753 6.964 6.488 5.804 6.194 6.238 Sr 0.004 < 0.002 0.002 0.006 0.004 0.004 Ba 0.003 0.001 0.003 0.003 0.003 0.003 Na 1.948 1.624 1.786 1.921 1.689 1.793 K 1.061 0.681 0.815 1.019 0.819 0.879 H- 0.091 0.023 0.138 0.090 0.046 0.068 H+ 2.015 1.517 1.513 2.518 2.780 2.072 C 1.953 1.638 1.599 1.770 1.128 1.622 C 1.946 1.647 1.595 1.749 1.119 1.612 O 58.132 58.373 58.435 57.983 58.336 58.252 57  Table 3.8: Whole rock trace element composition, Beluga occurrence. After Dzikowski (2013).  µg/g BA-1 BA-2 BA-3 BA-4 BA-5 Average Rb 80.2 47.1 56 71.2 55.9 62.08 Cs 3.5 1.9 2.5 3.2 2.1 2.64 Sr 305 131.5 230 368 423 291.5 Ba 64.9 38.5 49.7 51 69.5 54.72 Sc 11.8 13.8 12.3 9.9 11.2 11.8 V 218 263 236 221 247 237 Cr 110 130 100 100 100 108 Zr 196.0 248.0 212.0 198.5 220.0 215.0 Hf 6 8 7 7 7 7 Nb 4 3 4 3 3 3.4 Ta 0.5 0.5 0.5 <0.5 0.5 0.5 Mo <2 <2 <2 <2 <2 <2 W 1 1 1 2 3 ~ 2 Co 2.4 3.2 2.3 2.5 2.3 2.5 Ni 5 6 5 5 6 5.4 Cu 5 7 6 5 <5 ~ 5 Zn 32 35 31 29 34 32 Ag <1 <1 <1 <1 <1 <1 Ga 27 27 26 27 28 27 Tl <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 Sn 3 3 3 3 3 3 Pb <5 <5 <5 <5 <5 <0.5 Re 0.002 0.003 <0.002 0.003 <0.002 ~ 0.002 Se 3 9 4 4 3 4.6 Te <0.05 0.27 0.07 0.06 <0.05 ~ 0.1 Y 3.5 3.9 3.7 3.3 3.5 3.6 La 2.7 2.7 2.7 2.9 2.6 2.7 Ce 6.4 6.4 6.2 5.5 5.7 6.0 Pr 0.8 0.9 0.9 0.9 0.8 0.9 Nd 3.9 4.3 4.3 3.5 3.7 3.9 Sm 1.2 1.4 1.3 1.1 1.2 1.2 Eu 0.2 0.2 0.2 0.2 0.3 0.2 Gd 0.9 1 1 0.9 0.9 0.9 Tb 0.1 0.2 0.2 0.1 0.1 0.1 Dy 0.8 0.9 0.8 0.7 0.8 0.8 Ho 0.1 0.2 0.2 0.1 0.2 0.2 Er 0.4 0.5 0.4 0.4 0.4 0.4 Tm <0.1 0.1 0.1 <0.1 <0.1 ~ <0.1 Yb 0.4 0.5 0.4 0.4 0.4 0.4 Lu 0.1 0.1 0.1 <0.1 0.1 0.1 Th 1 1 1 1 1 1 U 1.7 1.8 2.1 1.7 1.7 1.8 Li 17.6 14.2 13.9 14.5 16.2 15.3 Be 1.84 2.01 1.93 1.78 1.84 1.88 B 60 60 90 100 80 78 F 1580 1050 1240 1390 1030 1260 Cl* 570 230 690 800 460 550 Cl** 490 270 620 810 510 540  *Neutron activation analysis, ** Specific ion electrode analysis   58  Table 3.9: Boron isotope composition of oxy-dravite from the Beluga occurrence.  δ11B (‰) σ (‰) 4.6  0.3 4.3  0.3 4.4  0.3 3.7  0.3 4.2  0.3 2.8  0.3 2.7  0.3 3.3  0.3 3.9  0.3 4.4  0.3 4.4  0.3 Average  3.9 0.7   59                      Table 3.10: U-Pb analytical data for zircon from the Beluga occurrence. Sample description1 A: N2, +180,t B: N2, +134,s,c C: N2, +134,s,c D: N2, +134,e,c E: N2, +134,e,c Wt. (mg) 0.090 0.033 0.011 0.032 0.023 U (ppm) 1233 608 632 776 1258 Pb2 (ppm) 355 169 170 229 385 206Pb/204Pb (meas.)3 3977 1294 1569 15970 1344 Tot. common Pb (pg) 513 277 76 30 418 % 208Pb2 1.5 2.0 2.1 1.6 3.0 206Pb/238U4 (± % 1s) 0.29734(0.19) 0.28611(0.14) 0.27582(0.17) 0.30505(0.07) 0.31143(0.11) 207Pb/235U4 (± % 1s) 4.4700(0.26) 4.3091(0.42) 4.1416(0.36) 4.5840(0.09) 4.6856(0.39) 207Pb/206Pb4 (± % 1s) 0.10903(0.14) 0.10923(0.35) 0.10890(0.28 0.10899(0.03) 0.10912(0.33) 206Pb/238U age (Ma; ± % 2 s) 1678.1(5.5) 1622.1(3.9) 1570.3(4.8) 1716.3(2.1) 1747.7(3.2) 207Pb/206Pb age (Ma; ± % 2 s) 1783.3(5.0) 1786.6(12.8) 1781.1(10.1) 1782.6(1.2) 1784.8(12.0) Discordance % 6.7 10.4 13.3 4.2 2.4 Th/U (calc.) 0.057 0.078 0.083 0.063 0.116 1N2 = non-magnetic at 2° side slope on Frantz magnetic separator; grain size given in microns; t = translucent; c = clear; s = stubby prismatic grains; e = slightly elongate prismatic grains. 2Radiogenic Pb; corrected for blank, initial common Pb, and spike. 3Corrected for spike and fractionation. 4Corrected for blank Pb and U, and common Pb.            60   Table 3.11: Estimated cooling rate and closure temperatures for phlogopite and muscovite at the Beluga occurrence using different determinations of activation energy (Ea) and frequency factor (D0) for phlogopite (Phl) and muscovite (Ms) and calculated in closure v1.2.   A 750 µM SPHERICAL RADIUS (APPROXIMATING A 500 µM CYLINDER RADIUS) WAS USED AS THE EFFECTIVE DIMENSION OF THE DIFFUSION DOMAIN  A ESTIMATE OF D0 FOR 5 KBAR B TWO POSSIBLE SETS OF EA AND D0 FROM A RE-EVALUATION OF GILETTI’S (1974) EXPERIMENTS * 0.01-10°C/MA COOLING RATE RANGE (VALUE IS A 9 DATA POINT AVERAGE) ** (PHL-A) – (MS-B)    Ms Ms Phl Phl Phl - Ms Phl - Ms  Phl Ms Ms data Phl data Ea (kJ/mol) D0 (cm2/s)  Ea (kJ/mol) D0 (cm2/s) ~ ΔTC (°C)* ~ Δt (Ma)** Estimated cooling rate (°C/Ma)  TC (°C) Approx. TC (°C) Approx. Robbins (1972), Hames & Bowring (1994) Giletti (1974) 180 4 × 10-4 242 242 67 124 0.54 400 330 Robbins (1972), Hames & Bowring (1994) Villa (2010)B 180 4 × 10-4 299.4 205 131 124 1.06 470 340 Robbins (1972), Hames & Bowring (1994) Villa (2010)B 180 4 × 10-4 359.2 93611 180 124 1.45 525 345 Harrison et al. (2009)A Giletti (1974) 268 20 242 242 -20 124 n/a   Harrison et al. (2009)A Villa (2010)B 268 20 299.4 205 44 124 0.35 455 410 Harrison et al. (2009)A Villa (2010)B 268 20 359.2 93611 93 124 0.75 515 425 61  3.8 Figures  Figure 3.1: Bedrock geology map of the True North Gems property with markers for scapolite, spinel, and corundum occurrences found during the 2006 mapping season. The most important mineralized areas are named. Small plutons consist of granite or ultramafic plugs. UTM zone 19 V (NAD83). Map courtesy of True North Gems Inc.  62     Figure 3.2: (A: top) Corundum (sapphire) gemstones from the Kimmirut occurrences. Left: colorless sapphire, Aqpik occurrence, 2.50 and 2.59 ct. Top centre: deep blue, extra fine sapphire, 1.17 ct, from the Beluga occurrence, and a heat-treated, rich blue 2.43 ct sapphire from Aqpik. Right: yellow sapphire, Beluga South occurrence, 1.09 ct and 1.47 ct. Photograph courtesy of True North Gems Inc. (B: lower left) lightly included, light blue sapphire (7.81 ct) from the Aqpik occurrence. Photograph by Brad Wilson. (C: lower right) dark blue corundum crystal, 36 × 4 mm, in calc-silicate rock, and a 1.17 ct sapphire gemstone from the Beluga occurrence. Photograph by Brad Wilson, courtesy of True North Gems Inc.   63   Figure 3.3: Contact between marble and sapphire-bearing calc-silicate rock at the Beluga occurrence. Note the variation in the distribution of light-colored and dark mineral assemblages within the calc-silicate pod, and the undulate nature of the contact with coarsely crystalline marble. The calc-silicate pod is surrounded by very coarse-grained marble, and no foliation is apparent in the outcrop.  64   Figure 3.4: Corundum-bearing calc-silicate rock in situ at the Beluga pit. Phlogopite-oligoclase (Phl + Pl) intergrowths occur near partly albitized scapolite (Ab + Scp) and albite, muscovite, and blue corundum (Ab + Ms + Crn). The corundum- and phlogopite-bearing assemblages contain calcite. Minor graphite (Cg) is present.  65   Figure 3.5: Diopside (Di) surrounded by phlogopite-oligoclase symplectite and coarser oriented intergrowths of phlogopite (Phl), oligoclase (Pl), and calcite (Cal). Beluga occurrence. Cross-polarized light. 66   Figure 3.6: Contact between coarse-grained scapolite (Scp), phlogopite-oligoclase-calcite (Phl-Pl-Cal), and the corundum-bearing zone. The latter zone contains idiomorphic corundum (Crn) with albite (Ab), calcite (Cal), and muscovite (Ms) of variable grain size. Dark zones of fine-grained alteration (Alt) consist of mixtures of the following minerals in variable abundance: albite, calcite, muscovite, analcime, prehnite, and thomsonite. Beluga occurrence. Plane polarized light. 67   Figure 3.7: Beluga calc-silicate rock under shortwave ultraviolet light showing the fluorescent scapolite (yellow; Scp), variable diopside-phlogopite-oligoclase assemblages (Di-Phl-Pl), albitized scapolite (Ab), and albite-corundum-muscovite assemblages (Ab-Crn-Ms). Minor calcite is present. Purple coloration is an artefact of the UV light source. 68   Figure 3.8: Thin scapolite rim between a grain of nepheline and calcite, which suggests that scapolite formed from the reaction of calcite and nepheline and post-dates the nepheline-bearing mineral assemblage. Bowhead occurrence. Cross-polarized light. 69    Figure 3.9: U-Pb concordia diagram for zircon recovered from the phlogopite-rich assemblage at the Beluga occurrence calculated using Model 1 of Ludwig (2003). Error ellipses represent 2σ. See Table 10 and text for U-Pb data and sample descriptions.       70  A (Phlogopite)  B (Phlogopite)  Figure 3.10 A-B: 40Ar/39Ar age spectra of phlogopite from the Beluga occurrence. Box heights are 2σ. Plateau steps are filled and rejected steps are open. After Dzikowski (2013).  71  C (Muscovite)  D (Muscovite)  Figure 3.10 C-D: 40Ar/39Ar age spectra of muscovite (Ms; C, D) from the Beluga occurrence. Box heights are 2σ. Plateau steps are filled and rejected steps are open. After Dzikowski (2013).  72   Figure 3.11A: Mineral reactions modeled in TWQ v1 (Berman 1988, 1991). Based on the paragenetic sequence of Bowhead and Beluga calc-silicate rock (present study) and metasediments on Aliguq Island (Butler 2007), the diopside (Di), nepheline (Ne), and K-feldspar (Kfs) assemblage is suggested to be possible at M1A metamorphic conditions but low XCO2 (see Fig. 3.11B). Alternatively, the formation of nepheline and diopside from albite (Ab) and dolomite (Dol) is consistent with regional P-T conditions but differs significantly from the paragenetic sequence at Beluga and Bowhead. The peak assemblage is partly replaced by phlogopite (Phl), calcite (Cal), and albite at conditions slightly below that of M2. The meionite (Me) – nepheline break-down reactions are probably overestimates since the measured scapolite compositions contain significant Na. The latter minerals break down to form corundum (Crn), calcite, and albite at higher temperature than a similar reaction forming Al-silicate, calcite, and albite. Regional P-T conditions after St-Onge et al. (2007). See text for a detailed discussion of mineral reactions and paragenetic sequence in the context of regional metamorphism.    73   Figure 3.11B: Mineral reactions modeled in TWQ (Berman 1988, 1991) modified after Butler (2007) at peak metamorphic pressure (8 kbar, St-Onge et al. 2007). The nepheline-bearing assemblage occurs at low XCO2 at the peak metamorphic temperature (810°C, St-Onge et al. 2007).  74   Figure 3.12: Comparison of Beluga calc-silicate rock major element composition with selected representative Lake Harbour Group metasediments (Thériault et al. 2001, Butler 2007), monzogranites (Butler 2007), lazurite/haüyne-bearing metaevaporites (Hogarth & Griffin 1978), averages for platform sediments (Carmichael 1989), and averages for syenite (Nockolds 1954).  75   Figure 3.13: Abundance of selected trace metals in Beluga calc-silicate rock compared to representative LHG metasediments (Thériault et al. 2001), sediments (Carmichael 1989, Moine et al. 1981 and references therein), and average granite (Carmichael 1989). 76    Figure 3.14: Abundance of V, Cr, and Ni relative to Al2O3. Line represents expected concentration of metals relative to alumina by varying amount of allochthonous detritus based on metal-Al2O3 ratios from a shale average (Carmichael 1989). Selected examples of metapelite and psammite from the Lake Harbour Group (Thériault et al. 2001) are included for comparison. 77    Figure 3.15: Chondrite-normalized (values of Taylor & McLennan 1985) REE trace element profile of Beluga calc-silicate rock compared to rocks from the Lake Harbour Group (*Butler 2007, **Thériault et al. 2001). After Dzikowski (2013).  78   Figure 3.16: Oxygen isotope composition of Beluga corundum compared to values from different types of gem corundum deposits (modified after Giuliani et al. 2014). Color in diamonds represent the colour of gem-quality corundum from different deposits (red = ruby, others = colored sapphire). Diagram courtesy of G. Giuliani.  79   Figure 3.17: Al-Mg-Ca diagram showing the distribution of the calc-silicate rock from the Beluga corundum deposit relative to domains for platform marls and clay-shale, evaporites, and meta-evaporites (modified after Moine et al. 1981) and in comparison to marbles (open symbols) and intercalated schists and gneisses (full symbols) from ruby-bearing marbles in central and south-east Asia (after Garnier et al. 2008). Diagram courtesy of G. Giuliani.   80   Figure 18: Boron isotope composition of Beluga oxy-dravite compared to tourmaline from different environments. Sources of data: (1) Swihart & Moore 1989; (2) Palmer 1991; (3) serendibite-rich calc-silicate rock interpreted to be metamorphosed illite layer deposited in a hypersaline environment (Grew et al. 1991); (4) serendibite-rich calc-silicate rock similar to the latter, but with both prograde and retrograde tourmaline (Belley et al. 2014). 81  Chapter 4. Metacarbonate-hosted spinel on Baffin Island: Insights into the origin of gem spinel and cobalt-blue spinel 4.1 Results summary  Fourteen spinel occurrences were sampled in the Lake Harbour Group (LHG), southern Baffin Island, Nunavut, Canada, and studied using a combination of petrography, whole rock geochemistry, microprobe analysis, and where possible, geochronology. Spinel at most occurrences is blue to violet in colour, and generally not of gem quality. Two spinel occurrences near Kimmirut contain facet- and cabochon-gem quality vivid blue, cobalt-enriched (0.03-0.07 wt.% CoO) spinel. The spinel mostly occurs in metasedimentary (sensu stricto) deposits, with the exception of two metasomatic occurrences in Markham Bay. All spinels occur in marble and calc-silicate/silicate-rich metacarbonate rocks. Minerals occurring with spinel as part of a stable assemblage include calcite, dolomite, phlogopite, pargasite, diopside, humite, forsterite, scapolite, anorthite, graphite, and pyrrhotite. Spinel formed under peak granulite facies metamorphic conditions. The spinel at two localities is partly replaced by retrograde corundum. Spinel-bearing metacarbonates are interpreted to have the following protoliths: (1) impure dolomite-bearing and dolomitic limestone; (2) dolomitic marl; and (3) evaporitic magnesitic marl. Spinel genesis in Mg-bearing metacarbonates is favored by: (1) the low abundance of silica relative to alumina, the primary control on whether spinel forms in most calc-silicate rocks; (2) low potassium activity limiting the formation of phlogopite and thus leaving Al available for spinel formation; (3) low XCO2 in marbles; and (4) insufficient quantities of Mg or dolomite reactant in diopsidite limiting Al incorporation into phlogopite to form spinel. The spatial distribution of Co enrichment at the cobalt-blue spinel occurrences is indicative of highly localized enrichment, with possibly only small scale (≤ 1 m) diffusion during 82  metamorphism. Cobalt and Ni are interpreted to have been enriched in the original sediment, or during diagenesis or low-grade metamorphism. Concentrations of Co (especially) and Ni are anomalously high (up to 29 µg/g, conservatively twice the expected cobalt concentration in a comparable metasediment), while concentrations of Fe, Mn, V, Cr, and Cu are much lower than expected; a chemical signature that speculatively could be caused by diagenetic processes prior to metamorphism. Pyrrhotite strongly partitions Fe relative to spinel, therefore an abundance of sulfide is expected to improve the attractiveness (therefore commercial value) of spinel by decreasing the amount of Fe incorporated into spinel (thus preventing overly dark colors). Marbles containing abundant metamorphosed dolomitic marl (magnesian calc-silicate) layers offer the best potential for gemstone discoveries in southern Baffin Island. Small, sometimes localized variations in whole rock Al/Si and K/Al, or fluid XCO2 could result in the occurrence of spinel. Other significant gemstone occurrences on Baffin Island (sapphire and lapis lazuli) are hosted in relatively similar rock types. 4.2 Chapter introduction Interest in spinel gemstones has increased significantly in the early 21st century (Pardieu et al. 2008); spinel with an intense red color (chromium chromophore) and vibrant cobalt-blue color (cobalt chromophore) are the most valuable. The enrichment of cobalt in concentrations sufficient to turn spinel bright blue has yet to be adequately explained. In the most detailed petrologic study on blue gem spinel to date, Chauviré et al. (2015) suggested that Co and Ni were enriched in cobalt-blue spinel from F- and Cl-bearing metamorphic fluids (partly derived from evaporites) sourcing these metals from evaporite or amphibolite. The enrichment of Cr in metacarbonates is better understood; research on ruby (Cr-bearing corundum) deposit models 83  suggest that the Cr enrichment is a result of Cr-rich sediment deposited coevally with carbonates and the concentration of Cr in organic matter (Garnier et al. 2008). At Revelstoke, British Columbia, Cr was enriched in gneiss layers that reacted with marble to form corundum (Dzikowski et al. 2014).  Very few areas in the world are producing gem spinel, and much of this spinel is recovered from eluvial or alluvial deposits. Perhaps because of this, the geology of spinel has yet to receive attention from geoscientists – in significant contrast to gem corundum geology (e.g., Giuliani et al. 2014). Moreover, the potential for gaining new insight into gem spinel genesis from small in situ occurrences that are not being mined for gem material (giving a much larger sample size) has been overlooked. This research will produce the first general gem spinel deposit model, based on 14 spinel occurrences in the Lake Harbour Group, Baffin Island, Nunavut, Canada (Table 4.1), containing violet, blue, and black spinel in addition to two cobalt-blue spinel occurrences and comparisons to similar, spinel-barren rocks. 4.3 Results 4.3.1 Petrography 4.3.1.1 Markham Bay A. Spinel Island This spinel occurrence is located on “Spinel Island” (Fig. 4.1; as named in this study, although it has no official name) just north of MacDonald Island, by the water in a diopsidite (calc-silicate) band at the contact between syenogranite pegmatite and dolomite-and-phlogopite-bearing calcite marble on the eastern tip of the island (N 63.69753° W 72.60063°; Fig. 4.2). A 55 meter portion of the calc-silicate band (1.5-2.5 m thick, strikes 089°, dip 80°) on the SE tip of the island contains spinel. The band has been truncated by erosion to the east, and re-appears on the 84  easternmost point of the island, where the diopsidite rarely contains small (< 1 cm) black spinel crystals. At the spinel occurrence, the calc-silicate band extends 30 meters to the west, where at its extremity, it occurs between lineated and foliated granitoid metaplutonic rocks (similar to that of Butler 2007 at the nearby Aliguq Island). The diopsidite in this area is medium-grained and devoid of spinel. South to north “lithosection” descriptions of the lithologies are presented in Tables 4.2A and B. At a point 40 meters south of the diopsidite band, only visible at low tide, coarse-grained monzogranite occurs south of the marble, and is surrounded by a metasomatic zone of diopsidite with large (3-20 cm) light brown phlogopite crystals and yellowish calcite. The monzogranite, which is different in composition to the pegmatite (more albite, presence of clinopyroxene and ilmenite, lack of titanite and rutile; sample 1A-MG), is coarse-grained, composed principally of albite, K-feldspar, quartz with subordinate clinopyroxene (intermediate between diopside and hedenbergite), ilmenite, trace apatite and zircon. Zircon, in addition to rare allanite and monazite were observed by use of a SEM and are generally < 20 µm in largest dimension. Marble west of the monzogranite contains small pods of white diopsidite and green talc/serpentine mixture. Spinel at this locality occurs as dark blue to black euhedral crystals up to 9 cm in size. The spinel is too dark to have value as a gemstone, but the locality has produced museum-quality mineral samples. Spinel crystals typically have equal development of octahedral and dodecahedral faces, and small trapezohedral {3 1 1} faces.  The spinel-bearing diopsidite is heterogeneous in both grain size and mineral abundance (Fig. 4.3A). It is very coarse to pegmatitic in calcite-rich zones, and medium-grained to very coarse in calcite-poor, diopside-dominant zones. Some portions of the rock contain over 30 volume % phlogopite. Euhedral spinel crystals averaging 1-3 cm (up to 9 cm) are most common 85  in pods of white calcite (5-35 cm), and are associated with euhedral, greenish-grey diopside prisms (1-3 cm diameter, 2-8 cm long) and dark brown phlogopite (up to 4 cm; Fig. 4.3B). Pink end-member anorthite is locally common and has been observed in association with spinel. Rare, euhedral thorianite crystals up to 1 mm occur with spinel (personal communication, B. Wilson; identified with qualitative EDS). The main assemblage was not subjected to significant retrograde alteration; only a negligible amount of pargasitization around the diopside and weak (15%) sericitization of the anorthite are apparent. The notable exception is a small zone (< 1 m) in the diopsidite band where the spinel is partly altered to corundum and other minerals, and where anhedral, greyish-green tremolite crystals up to 3 cm long are present. One microscopic inclusion of titanite was observed in the tremolite. Fractures in the partly altered spinel can contain corundum, sericite, probable chlorite, and/or gahnite. Very small amounts of zinc sulphide are associated with the corundum and chlorite alteration. The exterior of one spinel is penetrated by euhedral corundum crystals together with Mg-chlorite, which grades into a thin corona of margarite followed by zoisite/clinozoisite, and then tremolite (Fig. 4.3C; minerals identified by EDS). B. Unnamed Island Small quantities of black spinel occur in a metasomatic phlogopite- and calcite-bearing diopsidite on the edge of a small unnamed island located between “Spinel Island” and Aliguq Island (Fig. 4.1; N 63.70581° W 72.57546°). The majority (>95%) of the metasomatite is devoid of spinel, and instead contains abundant and large (1–15 cm, average 3 cm) euhedral, black amphibole crystals associated with dark brown phlogopite (≤10 cm) in calcite and diopside, especially bordering a network of calcite veins (Fig. 4.4). This calc-silicate unit occurs over an area of 10 by 3 meters on the edge of the island. Black, octahedral spinel crystals (1–30 mm) 86  with small dodecahedron faces occur in calcite just below the low tide mark. Microscopic chlorapatite (< 200 µm) was identified with semi-quantitative EDS. 4.3.1.2 Glencoe Island Multiple spinel occurrences are present on the north side of Glencoe Island (Fig. 4.5) in marble and calc-silicate rocks following a band of steeply-dipping marble in gneiss (the marble strikes approximately 104° in the NW part, and 124° in the SE part of the island). The geology of Glencoe Island was briefly described by Grice et al. (1982). The dominant rock type is biotite-quartz-feldspar gneiss with or without almandine with a band of calcite marble, up to 20 m wide, extending almost the full length of the north side of the island. K-feldspar-quartz pegmatites up to 2 m wide cross-cut the gneiss roughly perpendicular to the foliation. A.  Main occurrence Spinel occurs in calc-silicate pods in a 2-meter thick band of very coarse phlogopite-bearing serpentine marble (2A-M; N 63.08516° W 71.44985°). The serpentine occurs as pseudomorphs after what were presumably subhedral forsterite crystals, now entirely replaced. The serpentine marble contains trace graphite and, in one area, a very rich seam of graphite (10 × 80 cm) occurs with diopsidite. The band of marble is exposed on the side of a small embayment on the north side of the island and continues into the water (Fig. 4.6A). The contact between marble and siliceous host rocks strikes 110° and dips 68°. To the east end of the outcrop, the south contact of the marble is bound by psammite (2A-PS) and the north contact by dark green diopsidite (2A-DI). The psammite is principally composed of quartz, plagioclase (An80), and orthopyroxene (XMg ≈ 0.75) with subordinate pyrrhotite (< 2 vol. %) and graphite, and trace amounts of K-feldspar, titanite, and monazite. Parts of the psammite are very rich in quartz. Approximately 10-30 vol. % of the orthopyroxene is altered to a fine-grained hydrous Mg-87  silicate. The dark green diopsidite is dominantly composed of diopside with subordinate phlogopite. It contains 1-2 vol. % pyrite, graphite, and trace concentrations of pyrrhotite and chalcopyrite.  Spinel was first noted from this locality by Grice et al. (1982), who reported finding crystals up to 5 cm across. A 0.39 carat violet gemstone was faceted from a spinel from this location (Wilson 2014). Violet (as in sample 2A-SPL-1 and 2A-SPL-2) to dark greyish-violet (as in 2A-SPL-3) spinel crystals occur in calc-silicate pods ranging in size from 10 to 30 cm in maximum dimension. The pods tend to be unequally proportioned and are randomly oriented. The calc-silicate pods are composed of different assemblages in randomly distributed zones (Fig. 4.6B), such as: diopside, phlogopite-diopside, calcite-diopside-phlogopite-spinel (± pyrrhotite), and less commonly, anorthite-diopside (± calcite). Anorthite occurs in some of the spinel-bearing pods but was not observed in direct contact with spinel. Trace graphite occurs in all zones. Calcite-poor zones are medium- to very coarse-grained, and contain varying quantities of greenish-grey diopside and brown phlogopite. Very coarse-grained anorthite-diopside is less common. The calc-silicate rocks contain trace amounts of zircon, apatite, and Ba-rich K-feldspar. Spinel occurs only in and around calcite-rich zones, where grain sizes are generally between 4 and 25 mm. Spinel crystals are subhedral to euhedral, with equal development of octahedral and dodecahedral faces, but generally too fractured for gem cutting. Color ranges from pale violet to dark greyish-violet. Uncommon fractures in spinel are lined with chlorite and trace zinc sulphide. B. Marble beach occurrence Dark blue, euhedral spinel crystals up to 2 cm in maximum dimension occur in pods of silicate-bearing marble (< 25 cm) within a few blocks on marble over weathered marble outcrop 88  near the bay on the NE side of the island (N 63.07965° W 71.44720°). The spinels are transparent, but fractured and too dark to produce large gemstones. Their morphology resembles that of spinel at the main occurrence. In the silicate-bearing marble pods, crystals range between 5 mm and 3 cm in size, the majority of grains being 5-10 mm. The silicate-bearing pods are composed of calcite, phlogopite, grey diopside, and spinel in order of decreasing abundance.  C. Contact occurrence Dark greyish violet spinel occurs in a 1 meter thick steeply-dipping diopsidite between marble and gneiss and, locally, between marble and a 30-cm thick dyke of Kfs-Qtz pegmatite oriented sub-parallel to foliation (N 63.08924° W 71.47921°). Spinel has two modes of occurrence: (1) as ≤ 5 mm octahedra associated with very coarse-grained greenish grey diopside and dark brown phlogopite in white calcite pods (≤ 5 cm); and, (2) as elongated anhedral porphyroblasts up to 4 × 2 cm surrounded by a thin layer of medium-grained phlogopite in massive, medium-grained diopsidite. The latter spinels are elongated along the foliation, however the orientation of the phlogopite is variable.  D. Marble gulch occurrence White marble is exposed on a vertical wall in a gulch near the crest of a ridge (N 63.08957° W 71.48546°). The marble is composed of medium-sized grains with small quantities of forsterite, pargasite, trace graphite and apatite. A 1-3 cm layer of dolomite-bearing calcite marble contains greyish-violet octahedral spinel crystals up to 2 mm in size with 5 vol. % yellow forsterite (slightly serpentinized), local green euhedral amphibole (≤ 6 mm), and trace diopside. When associated with amphibole or diopside, the forsterite is strongly embayed by the former minerals (Fig. 4.7). Although the embayed grain boundaries could have formed by grain boundary migration, it is possible that the minerals do not form a stable assemblage and that 89  diopside and amphibole formed during retrograde metamorphism. No coronas or reaction textures were observed. 4.3.1.3 Kimmirut area Six spinel occurrences in the Kimmirut area were sampled for study (Fig. 4.8). Smaller occurrences of spinel (comparable to the Soper Lake Camp occurrence; see below) are common throughout the area. Two occurrences, Trailside and Qila, are located on True North Gems Inc.’s Beluga sapphire property and are significant for the exceptionally vivid blue color of the spinel. A. Soper River occurrence Violet spinel occurs in a 100 × 30 m area (N 62.964986° W 69.795387°; total 230 × 60 m total marble exposure) of dolomitic marble (3A-1) on the west side of Soper River, just east of the lapis lazuli occurrence (whole rock samples 3A-LAPIS-1, 3A-LAPIS-2) described by Hogarth (1971). The marble is coarse to very coarse-grained and is dominantly composed of carbonate (> 99 vol. %) consisting of white dolomite with 5 to 60 vol. % pale grey calcite. It strikes apparently NE and dips roughly 45°, but the coarse grain size, lack of distinct banding, and weathered exposure has obscured the attitude. Accessory minerals averaging 1-2 mm in maximum dimension (up to 6 mm) are much dispersed. From most to least abundant (relatively similar concentrations), they are: (1) white, rounded, fine-grained diopside and dolomite pseudomorphs after an unknown mineral; (2) lime green, euhedral pargasite; (3) violet euhedral spinel octahedra, less commonly with dodecahedral faces (Fig. 4.9); (4) diopside; (5) pyrrhotite; (6) rare graphite and phlogopite. Pargasite appears to consume grains of diopside, however unaltered diopside grains occur in close proximity (< 5 mm). Phlogopite is uncommon in the marble and was not observed in association with spinel. Wilson (2014) reported spinel crystals 90  up to 3 cm from this occurrence. The largest spinel observed by the author at Soper River measures 6 mm across. B. Soper Falls occurrence Spinel occurs in bright orange humite pods (up to 25 cm) in a low humite-bearing marble outcrop (Fig. 4.10; N 62.902903° W 69.839723°). The marble contiguous to the pods contains 20 % humite by volume and no spinel. The pods are aligned NNW; the marble is steeply-dipping, and the pod-bearing layer is traceable for 6 m, after which the weathered rock is covered in soil. Humite (≤ 5 vol. %) occurs in marble as far as 30 meters away from the occurrence. All rock in close vicinity (60 m) to the occurrence is marble, with the exception of a silicate rock included in the marble (K-feldspar, scapolite, diopside with subordinate quartz, calcite, trace apatite and Fe-oxide). The humitite consists of bright orange, very coarse-grained but highly fractured, fluorine-dominant humite (93-95 vol. %) with greyish-violet spinel octahedra (average 2 mm, up to 9 mm, 3-5 vol. %, generally associated with carbonate; Fig. 4.11A), white carbonate (magnesite >> dolomite, 2 vol. %), coarse pearly white talc (in 5-15 mm pods with carbonate, <1%, identified with XRPD), green pargasite (up to 2.5 × 1 cm), and uncommon pale grey diopside (partly replaced by tremolite) and pyrrhotite. Pargasite occurs in masses and thin veins cross-cutting humite rock. The pargasite appears to have formed at the expense of humite (Fig. 4.11B). Very thin rims of retrograde clinochlore occur around spinel. Dolomite is occasionally present around humite grains in magnesite and in fractures in magnesite; the dolomite appears to post-date the humite-spinel-magnesite stable assemblage (Fig. 4.11C). Talc has a similar grain size to magnesite, and does not appear to form at the expense of humite or magnesite (Fig. 4.11B). Talc is never in contact with spinel. 91  C. Soper Lake Camp marble Medium-grained white calcite marble is exposed on the edge of Soper Lake (N 62.871727° W 69.867925°) and contains 4 vol. % yellow forsterite, 1% pyrrhotite, <1% octahedral violet spinel, trace phlogopite, and graphite. The spinel is too small, fractured, and low in abundance to signal gem potential. One Ca-Mg amphibole crystal (EDS) was observed in contact with forsterite. Similar amphiboles are common in marble in this area. Rare titanium oxide, apatite, and one grain of zirconolite were identified with SEM-EDS.  D. Soper Lake mica mine Spinel occurs in a pegmatitic forsterite-phlogopite-carbonate rock exposed in a mica mine trench (N62.86540° W69.87154°; Fig. 4.12A). The trench is 21 m long and oriented N15°E. Samples of this silicate rock (3D-1 spinel-phlogopite-forsterite-calcite, 3D-2 forsterite-carbonate) were extracted for petrography and bulk rock geochemistry, however these are not representative of the average rock due to its extremely coarse-grained and heterogeneous nature. The mica ‘ore’ consists of very coarse-grained greyish-green forsterite (75-98 vol. %, estimate for entire rock unit) with dispersed brown phlogopite porphyroblasts (10–36 cm; ~ 5 vol. %) and randomly oriented pods of very coarse-grained white carbonate (dolomite and calcite, 2-10 vol. %, pods generally 5-15 cm in size) containing euhedral forsterite (≤ 3 cm) and, less commonly, dark greyish purple spinel octahedra up to 4.5 cm across (~ 0.5 vol. %; Fig. 4.12B), and trace graphite. Rare amphibole, pyrrhotite, and a 15 × 100 µm grain of probable dissakisite-(Ce) were identified using SEM-EDS. Some of the forsterite-rich rock locally has up to 60 vol. % carbonate. Forsterite is moderately serpentinized, between 10 and 30% by volume. Spinel gemstones have been cut from this occurrence (0.23 carat, very dark purple stone, Wilson 2014) 92  but the vast majority of spinel is not of gem quality, and no gem quality fragments were observed during field work for this study. To the east of the pegmatitic zone is marble. The unit contiguous to the band (3D-M1) is a coarse-grained forsterite-bearing marble with ~ 20 vol. % yellow forsterite, subordinate pale brownish-yellow phlogopite, trace dark violet spinel, and graphite. The medium- to coarse-grained marble (3D-3) several meters east of the pegmatitic zone consists of calcite (with dolomite exsolution) with 10-50% silicate, primarily phlogopite with subordinate forsterite, and trace graphite and pyrrhotite. E. Qila occurrence The Qila spinel occurrence is located approximately 1 km south of the Beluga sapphire pit. Incredibly vivid, cobalt-blue spinel occurs in a pargasite-rich calc-silicate pod and underlying marble (Fig. 4.13A). The calc-silicate rock forms the cap of a 1.2 m wide, 0.5 m high outcrop that has been eroded on all sides. Calcite-rich pargasite rock and spinel-bearing marble occur in the layers below, but are not exposed anywhere else in the vicinity of the exposure. The calc-silicate rock (3E-1) is very coarse-grained, consisting of pargasite, scapolite, phlogopite, carbonate, and spinel. In situ sampling of this zone was not permitted by the claim holder, but carbonate associated with spinel in float from the same unit consisted of dolomite. Subhedral spinel octahedra (2–10 mm across) are clustered in zones of higher concentration, sometimes with carbonate (Fig. 4.13B). The scapolite is meionite-dominant (Me69) with a thin exterior zone of sodic scapolite (Me29, EDS). Fractures in the scapolite are lined with Mg-chlorite. One grain of pargasite contained a microscopic diopside inclusion and rare pyrite grains up to 10 µm. 93  Some parts of the calc-silicate rock are underlain by a coarse-grained pargasite-calcite rock (3E-2), but the latter is very discontinuous. The principal mineral components of the coarse to very coarse-grained rock are pargasite (50 vol. %), calcite (45%; including minor dolomite exsolution), pale brownish-yellow phlogopite (5%), sky blue spinel (< 1%), and trace graphite, zircon, and apatite. One diopside grain in calcite, with a zone of pargasite (undulose contact) was observed in thin-section. Rarely, K-feldspar and Mg-Al-silicate (probable clinochlore) partly replace phlogopite. The marble underlying the pargasite-rich rocks consists of medium- to coarse-grained calcite (with dolomite exsolution) and approximately 6 volume % of other minerals, most commonly (in order of decreasing abundance): forsterite, pargasite, spinel, diopside, apatite, and phlogopite (3E-3-A). Isolated parts of the marble are richer in phlogopite relative to other silicates and contain only trace spinel (3E-3-B). The marble contains rare graphite and zircon. Forsterite grains commonly have diopside coronae, the exterior of which is sometimes replaced by pargasite (Fig. 4.14). Unaltered forsterite, partly altered forsterite, isolated diopside, and isolated amphibole occur in close proximity to each other (few mm) showing that equilibrium was not reached. Forsterite is commonly observed in contact with phlogopite. Diopside and amphibole were observed in contact with spinel. Since accessory minerals are widely dispersed in the marble, the paragenetic relationship between spinel and other silicate minerals cannot be ascertained. Spinel occurs as subhedral, vivid blue octahedra measuring 0.5-1 mm across. The blue color is not as intense as in the calc-silicate pod, but more vivid than in the calcite-rich pargasite rock. Two marbles were sampled within 30 meters of the Qila spinel occurrence. The first marble sampled (3E-M1) is coarse dolomitic marble with dispersed forsterite (Fo grains have 94  diopside coronae, and local F-rich pargasite, EDS), spinel, amphibole, diopside, graphite, and apatite.  A 6 mm purple spinel analyzed with EPMA has Co below the detection limit (< 0.03 wt. % CoO). The second marble (3E-M2) is coarse-grained, white calcite marble with less than 2 volume % of phlogopite, forsterite (variably serpentinized, in association with phlogopite, and having diopside coronae), and subhedral green diopside.  F. Trailside occurrence The Trailside spinel occurrence is located approximately 1 kilometer north of the Beluga sapphire pit. Cobalt-blue spinel occurs in two calc-silicate pods hosted in phlogopite-bearing marble. The main pod measures 3 × 2 meters. The calc-silicate pod contains irregularly distributed zones of different rock types, listed in order of decreasing abundance: diopside + phlogopite + carbonate + quartz (3F-4); sericite + phlogopite + clinochlore + calcite + spinel (3F-1); diopside + calcite; phlogopite + calcite; and phlogopite + calcite + spinel (3F-2).  The diopside-phlogopite-carbonate rock is principally composed of greenish-grey, very coarse-grained massive diopside with subordinate dolomite, phlogopite, quartz, and calcite. Trace apatite and zircon were observed in SEM. The phlogopite-calcite rock is coarse-grained with small amounts of fine-grained albite, and rare titanium oxide. The phlogopite is poikiloblastic (calcite inclusions) and randomly oriented. The main mineral components of spinel-bearing, mottled pale grey and brown calc-silicate rock (3F-1; Fig. 4.15) are sericite, phlogopite, calcite, clinochlore, albite, and spinel. Discontinuous, medium-brown phlogopite-rich bands occur throughout the rock. Cobalt-blue spinel is dispersed throughout the rock (≤ 1 vol. %) with slightly higher local concentrations. Two distinct mineral assemblages are apparent in thin-section (petrological microscope and 95  SEM): (1) a coarse-grained assemblage of brown phlogopite, yellowish calcite, and cobalt-blue spinel; and (2) a fine-grained assemblage of muscovite (sericite), clinochlore, calcite, albite, corundum, and pyrite. Titanium oxide was observed in SEM. Sericite-clinochlore mixtures (identified with X-ray powder diffraction) occur as pseudomorphs after subhedral crystals of a completely altered unknown mineral (an abundant, rock-forming component of the zone). Fine-grained alteration principally consists of sericite with up to 10% albite, and locally common clinochlore, with subordinate calcite. Muscovite and corundum are locally coarser-grained (crystals up to 0.6 mm). One grain of albite was observed in contact with phlogopite, where the phlogopite showed no alteration to muscovite or chlorite. Phlogopite and spinel are commonly surrounded by clinochlore. Spinel is altered to corundum and clinochlore in zones and along fractures (≤ 15 vol. %; Fig. 4.16). Pyrite associated with corundum-chlorite alteration is cobalt-bearing (see Mineral Compositions below). Vivid blue spinel occurs as subhedral octahedra averaging 4 mm across, rarely up to 2.7 cm. A small amount of spinel is of gem quality but due to the high degree of internal fracturing, the largest gemstone produced from this locality to date weighs 0.16 carats (Fig. 4.17; Wilson 2014). Cobalt-blue spinel also occurs in a medium-grained (1-3 mm), foliated phlogopite-calcite-spinel rock (3F-2; found as float above the outcrop; Fig. 4.18). The rock is composed, by volume, of pale beige carbonate (60%, calcite >> dolomite), pale brown phlogopite (35%), and vivid blue spinel (5%) with minor dolomite, scapolite (Me47, EDS), corundum and trace Fe-oxide and Ba-rich K-feldspar. Alteration of spinel to corundum is generally negligible (average < 5 vol. %).  The second spinel-bearing pod, located 24 meters SSW from the first, is smaller in size relative to the first pod and contains smaller spinel crystals (≤ 1 mm). Two samples of marble 96  were sampled in the vicinity of the spinel occurrence; samples 3F-M1 and 3F-M2. Marble 3F-M1 is calcite (with dolomite exsolution) marble with 5% phlogopite, 3% forsterite, trace diopside, and rare microscopic spinel and Fe-oxide. The forsterite is weakly serpentinized. One forsterite crystal is rimmed by a layer of diopside, and appears to have been replaced by diopside to some degree (Fig. 4.19). The second marble sample (3F-M2) is very coarse-grained white carbonate with ~ 1% phlogopite (2 mm crystals). Two spinel-free calc-silicate pods occur 48 meters south of the main spinel-bearing pod (3F-CS1, 3F-CS2, 3F-CS3). The rock ranges from a medium to very coarse grain size and the mineralogy consists primarily of diopside and phlogopite with subordinate dolomite, calcite, and quartz. In calc-silicate sample 3F-CS1, one crystal of calcic magnesian amphibole occurs with quartz and calcite as an inclusion in diopside. Sample 3F-CS2 also contains subordinate scapolite (Me55, EDS), trace plagioclase (An19, EDS), and rare zircon. 4.3.1.4 Waddell Bay occurrence A sample of coarse-grained pale orange calcite containing euhedral pale grey diopside, bluish-violet spinel, and dark brown phlogopite collected by Bradley Wilson was given to the authors for study. Wilson (2014) reported that blue and violet spinel crystals (combination of octahedral, dodecahedral, and trapezohedral faces) up to 3 cm occur with phlogopite and diopside in coarse-grained calcite marble. Wilson (2014) also reported that the largest faceted gem spinel from this location weighs 0.23 carats.  4.3.1.5 Unspecified occurrence on Hall Peninsula A sample of spinel-bearing marble from an unspecified location on the Hall Peninsula was given to the authors for study. The marble is a coarse-grained white calcite marble with 3% pale yellow forsterite, 1% dark greyish violet spinel (2 – 4 mm crystals), < 1% pale brown 97  phlogopite, and rare titanium oxide. Forsterite shows no alteration or reaction rims. Amphibole and calcite were observed as inclusions in spinel in thin-section. A phlogopite-calcite-diopside and a calcite-spinel inclusion occur in amphibole. The lack of obvious reaction textures and the dispersed nature of the silicates prevent definitive paragenetic sequence determination. 4.3.2 Mineral compositions Spinel and principal associated phases were analyzed with EPMA. Grains of each mineral were extracted from the same rock sample so that element concentrations in different minerals of a specific assemblage can be accurately compared.  Sample numbers refer to suites of minerals extracted from a single rock sample with the following exception: Sample 3A-1, which includes dolomite, calcite, pyrrhotite, amphibole, and diopside extracted from two rock samples, while a third sample (3A-2) contained spinel in contact with pargasite in calcite. Minerals at locality 3A (Soper River) were generally too dispersed to be extracted from the same rock samples. 4.3.2.1 Spinel Spinels in metacarbonates on Baffin Island are all Mg-dominant (spinel sensu stricto; Tables 4.3A-D), with a median XFe (Fe/[Mg+Fe]) of 0.07, a minimum of 0.03, and maximum of 0.18. Black spinel from the unnamed island in Markham Bay is richest in Fe (XFe = 0.18), followed by very dark blue spinel from Spinel Island (XFe = 0.14). Spinels with XFe ≈ 0.1 tend to be slightly dark but can be of gem quality (e.g., Glencoe Island main occurrence and Trailside). The spinels poorest in Fe, which are pale violet in color, are the pyrrhotite-associated Soper Camp spinel and the dolomite-hosted spinel in marble near Qila (XFe = 0.03). Vanadium concentrations are commonly below the detection limit (< 0.03 wt. % V2O3) with a median of 0.04 wt. % V2O3. Baffin Island spinels have a median of 0.06 wt. % Cr2O3, the highest value (0.14 wt. %) found in dark purple spinel from the Soper Lake mica mine. MnO 98  concentrations vary between below detection limit (< 0.03 wt. %) and 0.1 wt. % (median 0.045 wt. %). Spinels with the highest XFe (localities Markham Bay and two localities on Glencoe Island) also contain the highest Mn concentrations, however the values do not correlate perfectly. Cobalt is below detection limit (< 0.03 wt. % CoO) in all samples except at Qila and Trailside (0.03-0.07 wt. % CoO), where the spinel is vivid blue in color. Nickel was only detected in the Qila spinel (0.03-0.04 wt. % NiO, detection limit 0.03 wt. %). ZnO was measured in the majority of samples (detection limit 0.03 wt. %) with a median of 0.13 wt. %. Spinel from the main occurrence on Glencoe Island is particularly enriched in Zn (2.32-4.25 wt. % ZnO) relative to other Baffin Island spinels. A violet and a dark greyish-violet spinel from two different calc-silicate pods from the Glencoe main occurrence had slightly different major element concentrations: 2.32 and 4.25 wt. % ZnO, < 0.02 and 0.04 wt. % V2O3, but relatively similar FeO (4.88 vs. 4.23 wt. %), respectively. At the Soper Lake mica mine, spinel in phlogopite “ore” is richer in chromium relative to spinel in the contiguous Fo-Spl marble (0.14 and 0.05 wt. %, respectively), while the spinel in marble contains more ZnO (0.18 in the marble vs 0.12 wt. %). Three vivid blue spinel samples were extracted from different lithologies at Qila: calc-silicate rock, pargasite-calcite rock, and spinel-bearing marble. Spinel from the calc-silicate rock is richer in ZnO (0.16 wt. %), Fe (XFe = 0.07), CoO (0.07 wt. %), and NiO (0.04 wt. %) relative to spinel from the other rock units (0.05 wt. % ZnO, XFe = 0.05, 0.03 wt. % CoO, ≤ 0.03 wt. % NiO). A pale violet spinel recovered from dolomitic marble near the Qila outcrop contains 0.09 wt. % ZnO, a lower XFe (0.03), and no detectable NiO or CoO (< 0.03 wt. %). This pale violet spinel is much richer in V2O3 (0.04 wt. %) than the cobalt blue spinel at Qila (< 0.02 wt. %). 99  Cobalt-blue spinels from the Trailside occurrence are all Cr-, V-, and Ni-poor (below detection limits), contain ~ 5 wt. % FeO, and 0.06 wt. % CoO. Spinel in the silicate rock contains roughly half (0.05 wt. %) the amount of ZnO compared to spinel in the calcite-phlogopite-spinel rock (0.13 wt. %). 4.3.2.2 Forsterite Olivines in the Baffin Island spinel-bearing metacarbonate samples are all end-member forsterite (Fo97 – Fo99; Table 4.4). Forsterite associated with cobalt-bearing, vivid blue spinel in the Qila marble contains up to 0.03 wt. % CoO. Greenish and grey forsterite from the same sample of Qila spinel-bearing marble vary slightly in FeO (2.39 and 2.28 wt. %) and MnO (0.05 and 0.07 wt. %) concentrations, respectively.  4.3.2.3 Diopside Diopside associated with spinel (Table 4.5A-B) is very Na-poor (< 0.01 apfu), with the exception of low concentrations at Qila (0.026 apfu) and Waddell Bay (0.043 apfu). They generally contain 0.116-0.235 Al pfu except for Al-poor diopside from Soper Falls, Soper River (~ 0.03 apfu), and in the diopside-dolomite pseudomorph at Soper River (0.003 Al pfu). All diopsides are relatively Fe-poor; XMg ranges between 0.95 and 0.99. The most Fe-rich diopside originates from localities with the darkest colored, most Fe-rich spinel.  4.3.2.4 Amphibole All amphiboles are pargasite with the exception of tremolite replacing diopside at Spinel Island and Soper Falls (Tables 4.6A-C). Tremolite from Spinel Island is the most Fe-rich (XMg = Mg/(Mg+Fe) = 0.93). Pargasite from the calc-silicate rock at Qila has XMg = 0.97 while all other amphiboles have XMg = 0.98 or 0.99. 100  Pargasite is generally TiO2-poor (< 1 wt. %) with the exception of that at Glencoe Island marble gulch, Soper Lake Camp, and the Hall Peninsula marble occurrence, which have TiO2 concentrations between 1.69 – 2.64 wt. %. The A-site is Na-dominant with XNa = Na/(Na+K+Ca) ranging between 0.55 and 0.93 (lowest at Glencoe gulch, highest at Soper Falls). Pargasite from Qila is particularly rich in K2O (up to 2.23 wt. %). Chlorine contents vary between 0.04 and 0.18 wt. %, and most fluorine concentrations are below detection limit (< 0.12 wt. %). Pargasite, which formed at the expense of F-rich humite, contains 0.42 wt. % F. Barium is generally below detection limit (< 0.06 wt. %) but up to 0.21 wt. % at Glencoe marble gulch. Pargasite in marble at the Glencoe gulch occurrence contains small zones richer in TiO2 (2.64 vs. 1.69 wt. %), poorer in Fe2O3 (0.43 vs. 0.60 wt. %), and poorer in K2O (0.44 vs. 1.05 wt. %). Pargasite in Soper River marble varies considerably in TiO2 (0.23-0.58 wt. %) and Cl (0.08-0.18 wt. %) concentrations, with moderate variation in Na2O (2.61-3.00 wt. %). At Qila, pargasite in the calc-silicate rock is richer in the following components relative to that in the pargasite-calcite rock: BaO (0.12 vs. 0.07 wt. %), K2O (2.23 vs. 2.02 wt. %), and F (0.20 vs. < 0.12 wt. %). Pargasite in the pargasite-calcite rock contains more Na2O (2.74 wt. %) than pargasite in the calc-silicate unit (2.04 wt. %). 4.3.2.5 Phlogopite Phlogopites associated with spinel in metacarbonates on Baffin Island contain an average of 0.62 wt. % TiO2 with a range of 0.09 – 1.60 wt. % (Tables 4.7A-E). FeO contents are 1.09 wt. % on average, ranging from 0.45 to 2.30 wt. %. Vanadium is below detection limit with the exception of the phlogopite associated with dark spinel at the Glencoe main occurrence (0.04 wt. %). The median BaO content is 0.83 wt. %; BaO values are above detection limits in all samples, 101  the lowest value being at Trailside (0.10 wt. %) and with Ba-rich outliers at Glencoe gulch and in the Hall Peninsula marble sample (4.04 – 6.64 wt. % BaO). Fluorine concentrations vary between below detection limit (< 0.12 wt. %) and 1.58 wt. %. Phlogopite richest in F is from the Qila calc-silicate rock (1.58 wt. %), Soper Lake mica mine (1.52 and 1.32 wt. %), and Spinel Island (1.22 wt. %). At the Glencoe main occurrence, phlogopite associated with violet spinel differs in composition from that associated with the dark greyish violet spinel: they contain, respectively, 0.92 and 0.28 wt. % TiO2, 1.51 and 1.27 wt. % FeO, and 2.21 and 1.40 wt. % BaO. At the Soper Lake mica mine, phlogopite in the Fo-Spl marble is richer in FeO, Na2O, BaO, and poorer in TiO2 and F compared to phlogopite in the mica “ore.” Phlogopite occurs in the calc-silicate rock, the pargasite-calcite rock, and spinel-bearing marble at the Qila spinel occurrence. The phlogopites have the following compositional ranges, with the richest sample mentioned in brackets: 0.29-0.60 wt. % TiO2 (calc-silicate), 0.51-0.80 wt. % FeO (calc-silicate), 0.51-0.95 wt. % BaO (Pargasite-Calcite rock), 0.11-0.72 wt. % Na2O (Prg-Cal rock), 0.07-0.14 wt. % Cl (Pargasite-Calcite rock), and 0.42-1.58 wt. % F (calc-silicate). In the previously listed elements, the composition of phlogopite from the marble is intermediate between the calc-silicate and pargasite-calcite units with the exception of fluorine, where it is the least abundant. Two phlogopite grains from the Trailside spinel-bearing silicate rock were analyzed. They are relatively similar, with the exception of different TiO2 (0.09 – 0.32 wt. %), BaO (0.10-0.29 wt. %), and FeO (1.33-1.25 wt. %) contents. Phlogopite from the calcite-phlogopite-Spinel rock is, relative to that from the silicate rock, fairly comparable in TiO2 and FeO concentrations. 102  Phlogopite from the calcite-phlogopite-spinel unit is richer in BaO (0.35 wt. %), Na2O (0.44 wt. %), and Cl (0.07 vs. ≤ 0.02 wt. %). In spinel-bearing marble from the Hall Peninsula, the Ba-rich (4.30 wt. % BaO) phlogopite contains slightly Ba-enriched zones (average 6.64 wt. % BaO) that correlate with slightly richer TiO2 content. 4.3.2.6 Other Silicates and Oxides The Soper Falls humite is F-dominant, containing 4.70 wt. % fluorine. It contains 1.18 wt. % Fe, and 0.43 wt. % TiO2 (Table 4.8). The high total (103.07 wt. %) may be influenced by lower precision on F measurements. A fragment of scapolite from the spinel-bearing calc-silicate pod at Qila is predominantly meionite (Me69; Table 4.9) but thin exterior rims of marialite (Me29) were identified using EDS. The measured composition of scapolite (EPMA data) is low since CO was not measured or accounted for. Muscovite, part of a retrograde mineral assemblage at Trailside, contains small amounts of Na, Ca, Mg, and Ba (Table 4.10). Corundum replacing spinel as part of the same assemblage has the following composition as determined with EPMA (weight %): 100.76 Al2O3, 0.07 FeO, and 0.03 Cr2O3. 4.3.2.7 Calcite and Dolomite Calcite (Table 4.11A-B) is the most common carbonate associated with spinel on Baffin Island. Dolomite commonly occurs as exsolution in calcite. Only Soper River marble, a spinel-bearing marble near Qila, and the Qila calc-silicate rock contain significant quantities of dolomite as part of the stable assemblage (i.e., not exsolved dolomite within calcite; Table 103  4.12A-B). Small amounts of magnesite and dolomite, in addition to calcite, occur in humitite at Soper Falls. Calcite contains from < 0.03 to 0.27 wt. % FeO, far less than dolomite (0.18 – 0.74 wt. %). Rock-forming abundances of dolomite and calcite occur together at Soper River; the dolomite is significantly richer in FeO relative to calcite (0.23 vs. 0.03 wt. %) but they have identical MnO concentrations (0.04 wt. %). MgO concentrations in calcite generally vary between 0.35 and 3.45 wt. % with the exception of exceptionally Mg-poor calcite at Spinel Island (0.03 wt. %). Calcite in the high range of MgO content has very fine dolomite exsolution, which is reflected in the EPMA data. 4.3.2.8 Pyrrhotite and Pyrite Pyrite containing 1.79 wt. % cobalt (Table 4.13) occurs with corundum in altered spinel at the Trailside occurrence (the slightly low total weight % may be due to error since the grain was small, ca. 10 µm in maximum dimension). Pyrrhotite is the only sulphide mineral occurring with spinel as part of a stable assemblage. Pyrrhotite occurs in spinel-bearing calc-silicate pods at the Glencoe main occurrence, in spinel-bearing marble at Soper River and Soper Lake camp, and in the humitite at Soper Falls. Cobalt concentrations range from 0.06 to 0.11 wt. %; Ni varies from below detection limit (<0.03 wt. %) to 0.21 wt. %. 4.3.3 Whole-rock compositions Whole rock major element (Tables 4.14A-C) and trace element (Tables 4.15A-D) compositions were obtained for selected rock samples. Of the number of samples analyzed, the results of six are considered somewhat qualitative due to limited sample size (very coarse-grained rock or extremely dispersed minerals not adequately represented in the collected samples): (1A-SG) Spinel Island syenogranite; (1A-SPL) Spinel Island spinel-bearing diopsidite; 104  (3D-2) Soper Lake mine rock from the main pit, which contains forsterite and carbonate, but no phlogopite or spinel; (3A-1) marble from Soper River consisting of two fairly representative marble samples from the spinel-bearing unit; (3E-1) Qila calc-silicate rock sample lacking the spinel and carbonate found in this rock unit; and (3F-1) Trailside silicate-rich rock with spinel that did not contain significant spinel or phlogopite (relative to the darker bands in rock, see petrography). The low total in humitite is due to the large amount of fluorine not included in the total. Similarly, Cl was not included in the total for lapis lazuli. Among others, the elements Sc, V, Cr, Mn, Fe, Co, Ni, and Zn are of particular interest to the study of gem spinel due to their abundance in the mineral and due to the role of Cr, Fe, Co, and V as chromophores which directly control the beauty, and therefore market value of gem spinel. Selected light elements (Li, B, Cl, F) are also discussed below. 4.3.3.1 Lithium All marble samples and the Soper River lapis lazuli contain Li concentrations below detection limit (< 10 µg/g). Spinel-bearing rock at Spinel Island is relatively poor in Li (10 µg/g), however phlogopite is under-represented in the sample. Spinel-bearing calc-silicate rock and psammite at the Glencoe main occurrence contain between 10 and 30 µg/g Li. The Co-enriched calc-silicate pod at Qila contains 30 µg/g Li, while the spinel-bearing rock at Trailside is comparatively rich in Li (30-50 µg/g). The Co-rich Trailside diopsidite contains low Li (< 10 µg/g), and the spinel-free calc-silicate pods in the vicinity of Trailside contain 10 µg/g Li.  4.3.3.2 Boron, Chlorine, Fluorine These elements were measured only in selected samples: Glencoe main spinel-bearing calc-silicate, Soper Lake Forsterite-Spinel-bearing marble, Soper Falls humitite, Soper River 105  spinel-bearing marble (B only), Qila Pargasite-Calcite rock, Trailside spinel-bearing silicate rock, Trailside area marble, and Soper River lapis lazuli. The marbles and lapis lazuli are poorest in boron, 13-30 µg/g, with the exception of B-rich (129 µg/g) Forsterite-Spinel marble contiguous to the pegmatitic forsterite-phlogopite-carbonate-spinel rock at the Soper Lake mica mine. Spinel-bearing calc-silicate rock at the Glencoe main occurrence contains 21 µg/g B, much less than at Qila (95 µg/g), Trailside (196 µg/g), and Soper Falls (304 µg/g). Lapis lazuli and the Soper Lake mine Forsterite-Spinel marble are poorest in F (560 and 510 µg/g). Spinel-bearing rock at the Glencoe main occurrence contains 1070 µg/g F. The Trailside cobalt-blue spinel bearing silicate rock and the phlogopite-bearing marble from the Trailside area contain similar F concentrations (2320 and 1970 µg/g). Pargasite-rich Qila calc-silicate rock is, unsurprisingly, much richer in F than the latter rocks (6960 µg/g). Soper Falls spinel-bearing humitite, which is dominantly composed of F-dominant humite, had a fluorine concentration exceeding the upper limit of quantification (> 2 wt. %). Chlorine is least abundant in humitite and marbles (250-410 µg/g). The spinel-bearing silicate rock at Trailside contains 680 µg/g Cl, while the spinel-bearing rock at Glencoe main is slightly richer in the halogen (820 µg/g). The Qila pargasite-rich rock is, unsurprisingly, also richer in Cl with 1160 µg/g, and the Soper River lapis lazuli is richest in the element (3170 µg/g). 4.3.3.3 Chromium and Vanadium LHG marble samples contain an average of 9 µg/g vanadium (5-17 µg/g range). In comparison, calc-silicate rocks, humitite, and lapis lazuli contain 22-53 µg/g V. Spinel-bearing rock at Spinel Island is relatively poor in V compared to the other calc-silicate rocks (14 µg/g). 106  Forsterite-carbonate rock from the Soper Lake mica mine contains very low V (5 µg/g), probably owing to the lack of spinel or phlogopite in the sample and thus, the value is likely not representative of the rock unit. The psammite at Glencoe Island contains 29 µg/g V, within range of the calc-silicate rocks, and the sulphide-rich diopsidite at the same location is exceptionally V-rich (457 µg/g). At the Qila cobalt-blue spinel occurrence, amphibole-rich units are richer in V (13 µg/g) relative to the marbles (≤ 5 µg/g). At the Trailside cobalt-blue spinel occurrence, phlogopite-calcite-spinel rock (3F-2) and diopsidite (3F-4) contain 17 and 16 µg/g V, respectively. Trailside sample 3F-1 is poorer in V (7 µg/g), likely a reflection of the low abundance of spinel and phlogopite in this sample relative to this rock unit. Chromium concentrations in LHG marble, including those containing cobalt-blue spinel, are ~10 µg/g. Two spinel-bearing calc-silicate pods at Glencoe Island were found to contain 60 and 100 µg/g Cr. Spinel-bearing humitite is intermediate at 40 µg/g. Soper River lapis lazuli contains 30-50 µg/g Cr. Calc-silicate pods in the vicinity of the Trailside occurrence contain 20-30 µg/g Cr, while rocks at the Trailside occurrence are slightly poorer in the element (10-20 µg/g). The main calc-silicate pod at Qila (40 µg/g) is richer in Cr relative to other rocks at this occurrence (10 µg/g). As with vanadium, the low (10 µg/g) Cr concentrations in the Soper Lake mine forsterite-carbonate rock are probably not representative of the entire rock unit. 4.3.3.4 Manganese and Iron Marbles and silicate-rich metacarbonates range in MnO content from 0.03-0.06 wt. % and 0.03-0.08 wt. %, respectively. Lapis lazuli is, in comparison, Mn-poor with 0.02 wt. % MnO. Glencoe Island sulphide-bearing psammite and diopsidite are both intermediate, with 0.04 wt. % MnO. Co-rich (9-29 µg/g) rocks at Qila and Trailside contain low to moderate (0.02-0.04 wt. %) MnO concentrations relative to other metacarbonates. 107  Iron concentrations are expressed as total Fe = Fe2O3. The Glencoe main occurrence diopsidite and psammite are, by far, the richest in Fe (5.36 and 4.89 wt. %, respectively). Silicate-rich metacarbonates are poorer in iron, ranging from 1.11-1.88 wt. % Fe2O3 with higher concentrations in the forsterite-carbonate rock at the Soper Lake mine (2.58 wt. %). Lapis lazuli is Fe-poor for a silicate-rich metacarbonate (~0.4 wt. %). Marbles contain less iron than calc-silicates, 0.31-0.94 wt. %. The Co-enriched rocks at Qila are relatively Fe-poor (0.33-0.82 wt. % Fe2O3) while those at Trailside are closer to the range for other metacarbonates in the LHG (0.44-1.14 wt. %). 4.3.3.5 Cobalt Cobalt concentrations in the LHG metacarbonates are generally low: < 1 µg/g in lapis lazuli from Soper River; < 1 to 3 µg/g in marbles and impure marbles; and 2 to 6 µg/g in calc-silicate rocks. At the Glencoe Island main occurrence, sulphide-rich psammite and diopsidite are richer in Co, containing 7 and 16 µg/g respectively. Calc-silicate pods devoid of spinel in the vicinity of the Trailside occurrence are in the high range of cobalt concentrations for a calc-silicate rock (considering the lack of sulphides); 5 and 8 µg/g. Rocks containing cobalt-blue spinel, or spinel-free zones near these rocks (i.e., Trailside diopsidite and Qila phlogopite-bearing, spinel-poor marble) are comparatively rich in cobalt; 16 µg/g on average, with a range of 9-29 µg/g. The cobalt concentration in the main calc-silicate pod sample (3E-1; pargasite, scapolite) is likely an underestimate, since a sample with spinel (the mineral in the assemblage that most strongly uptakes Co, as shown in the EPMA data) could not be taken due to sampling restrictions on the claim property. Similarly, the Trailside silicate rock (3F-1) was spinel-poor and did not contain a representative amount of the darker bands richer in phlogopite and spinel. 108  4.3.3.6 Nickel Nickel concentrations in the LHG marble samples vary between < 1 and 4 µg/g. Concentrations of Ni in the lapis lazuli are below detection limit (< 1 µg/g). In silicate-rich metacarbonates, Ni concentrations vary between 2 and 23 µg/g with the exception of Soper Falls humitite (< 1 µg/g). The highest Ni concentration measured in the calc-silicates (23 µg/g) is in a pyrrhotite- and spinel-bearing rock from the Glencoe main occurrence. Nickel is fairly enriched in the Glencoe Island sulphide-bearing psammite (35 µg/g) and diopsidite (149 µg/g). Relative to most metacarbonates, rocks containing cobalt-blue spinel are richer in Ni, containing 22 µg/g on average (10-47 µg/g). 4.3.3.7 Copper The LHG marbles sampled contain < 1 to 6 µg/g Cu, with the higher concentrations being in pyrrhotite-bearing samples. Copper was not detected in lapis lazuli (< 1 µg/g). In silicate-rich metacarbonates, copper concentrations range from 1 to 17 µg/g and, once again, the highest concentration is in sulphide-bearing rock (Glencoe Island 2A-SPL-2). Glencoe sulphide-bearing psammite and diopsidite are the rocks richest in copper with 62 and 55 µg/g, respectively. Rocks associated with cobalt-blue spinel contain negligible Cu (≤ 2 µg/g). 4.3.3.8 Zinc Zinc is most abundant in the Glencoe main occurrence, where it occurs in concentrations between 230 and 308 µg/g in the spinel-bearing pods, serpentine marble, and sulphide-rich diopsidite; less so in the psammite (67 µg/g). At Spinel Island in Markham Bay, phlogopite-bearing marble is Zn-poor (< 2 µg/g) while the spinel-bearing rock contains 98 µg/g. Syenogranite pegmatite in contact with the spinel zone on Spinel Island in Markham Bay contains 7 µg/g Zn, and the nearby monzogranite, 14 µg/g. 109  Other LHG silicate-rich metacarbonates contain between 4 and 14 µg/g Zn. Marbles contain up to 5 µg/g. Rocks at the Qila cobalt-blue spinel occurrence are Zn-poor (≤ 2 µg/g) while the rock at Trailside varies considerably (two samples ≤ 3 µg/g, one spinel-rich sample containing 15 µg/g Zn). As in the case of cobalt concentrations, the low abundance of Zn in sample 3F-1 may reflect sampling bias. 4.3.3.9 Rare earth elements  Whole rock samples of Lake Harbour Group marbles and calc-silicate rocks have a similar chondrite-normalized REE profile to that for LHG metasediments reported by Thériault et al. (2001), characterized by fractionated LREE, relatively flat HREE profile, and a negative Eu anomaly. The profile for Glencoe Island psammite is similar, but with no Eu anomaly. Qila calc-silicate rock and diopsidite at and near the Trailside occurrence have a flat REE profile with negative Eu anomaly. The chondrite-normalized REE profile of Spinel Island monzogranite is very similar to that of the syenogranite pegmatite; both are downward sloping with increasing atomic number and have a positive Eu anomaly. 4.3.4 Pseudo-sections T-X(CO2) pseudo-sections at peak metamorphic conditions (P = 8 kbar, St-Onge et al. 2007) were calculated using Perple_X (Figs. 4.20 to 4.23) for the following samples: (2A-SPL-2) diopside-phlogopite-spinel-carbonate rock at the Glencoe Island main occurrence; (3A-1) dolomite-calcite marble with subordinate spinel, diopside-dolomite replacements after an unidentified mineral, pargasite, and phlogopite from the Soper River occurrence; (3E-3-A) calcite-dolomite marble containing forsterite, diopside, pargasite, spinel, phlogopite, and apatite from the Qila occurrence; and (3F-2) phlogopite-calcite-dolomite-spinel rock with trace scapolite from the Trailside occurrence. The following model systems were used: (2A-SPL-2) K2O-CaO-110  MgO-FeO-SiO2-Al2O3-H2O-CO2; (3A-1 and 3E-3-A) Na2O-K2O-CaO-MgO-FeO-SiO2-Al2O3-H2O-CO2; and (3F-2) Na2O-K2O-CaO-MgO-FeO-SiO2-TiO2-Al2O3-H2O-CO2. Models that include Na result in the prediction that accessory aspidolite (Na end-member of phlogopite) is associated with phlogopite under certain conditions. The biotite solution model does not account for the substitution of Na for K, and thus the result is interpreted to reflect a small increase in phlogopite Na content. This is considered plausible because even in cases of extreme Na-enrichment in such micas, where the mineral species aspidolite is formed, the crystals contain mixed layers of variable Na and K content (e.g., mixed aspidolite-phlogopite crystals associated with spinel in marble, Belley et al. 2016; mixed aspidolite-phlogopite-paragonite in ruby-bearing marbles, Garnier et al. 2004). No solution model was used for Mg incorporation in calcite, which may affect the positioning of phase boundaries separating calcite vs. calcite-dolomite bearing assemblages. In calc-silicate sample 2A-SPL-1 (Glencoe Island main occurrence; Fig. 4.20), spinel is stable at all XCO2 at peak metamorphic conditions. The mineral assemblage noted in petrography is calculated to occur at XCO2 between ca.0.45 and 0.85. At peak metamorphic conditions in the T-X pseudo-section for Soper River marble sample 3A-1 (Fig. 4.21), spinel occurs in the range ca. 0.09 ≤ XCO2 ≤ 0.57. Spinel, forsterite, and pargasite occur as part of a stable assemblage without diopside between XCO2 = 0.3 and 0.5. The forsterite-out occurs at lower temperature, where the assemblage consists of pargasite, diopside, phlogopite, calcite, and dolomite. Spinel-bearing marble from Qila (sample 3E-3-A) produced a similar pseudo-section to that of Soper River marble; spinel is only stable at lower XCO2 (~ 0.07 to 0.48, Fig. 4.22). At peak metamorphic conditions and with increasing XCO2, the magnesian silicate assemblage 111  changes from containing forsterite, to forsterite-pargasite, to forsterite-pargasite-diopside, pargasite-diopside, and diopside. In the phlogopite-calcite-spinel rock at Trailside (3F-2), spinel is expected to occur throughout the XCO2 range during peak metamorphism (Fig. 4.23, excluding partial melting at very low XCO2). Trace scapolite occurs in this sample but was not observed in contact with spinel. Scapolite and spinel occur as part of a stable assemblage together with phlogopite, calcite, dolomite, and nepheline at XCO2 ~ 0.95. At lower temperature and high XCO2, one assemblage consists of scapolite-phlogopite-calcite-dolomite. A spinel-free corundum-bearing assemblage also occurs at lower temperature. 4.4 Calculations 4.4.1 Sedimentary protolith composition estimation: Method and underlying assumptions Classifying various LHG metasedimentary carbonates and impure carbonates can provide insights into protoliths that favor spinel genesis under high-grade metamorphic conditions. Metasedimentary rocks that are not the result of reaction with metasomatic fluids or plutonic rocks (i.e., Markham Bay localities) are likely representative of the compositions of the sedimentary protoliths, although it should be reminded that sampling bias (discussed for specific samples in section 4.3.3; some examples have been excluded from composition estimation for this reason) and major element mobility during metamorphism will affect these results. Under the assumption that whole rock compositions are representative of the original rock, we qualitatively estimate the sedimentary composition of the protolith by: (1) calculating an estimate of siliciclastic sediment abundance using Al/Si (Fig. 4.24) to estimate mixing of “ideal” sandstone, shale, and kaolinite-rich claystone (for Al/Si values considerably higher than typical shale or clay); (2) using this estimate to calculate the expected CaO and MgO content of the 112  siliciclastic fraction; (3) calculating an estimate of original carbonate CaO and MgO by subtracting the former values from the totals; (4) calculating the estimated relative abundance of calcite, dolomite, and magnesite carbonate species in the protolith; and (5) re-calculation of the sediment weight percentages using the original protolith CO2 content estimate and assuming that the shale/clay fractions contain 5 wt. % H2O, which will account for mass loss during metamorphic devolatilization reactions. All non-siliciclastic CaO and MgO, in addition to CO2 are considered part of the carbonate fraction. Siliciclastic CaO and MgO are assigned to their respective sediment fractions. Other elements, especially minor elements, are divided proportionally between the siliciclastic components. Lapis lazuli from Soper River is exceptionally Na-enriched – far more than would be possible in shale or clay. While Na is very mobile and could have been introduced during metamorphism, we attribute Na to halite in the original protolith calculation and add Cl, based on the identification of Soper River lapis lazuli as a meta-evaporite by Hogarth & Griffin (1978). Results are presented in Tables 4.16A-D. 4.4.2 Estimate of “expected” minor metal content Comparing metasediment whole rock trace element abundances (or relative abundances, e.g., metal/Al) to that of typical sedimentary rocks is a useful tool in assessing the geochemical properties of the metasedimentary rock, which is expected to reflect original protolith composition and potential chemical changes during diagenesis and metamorphism. Variations in some elements can imply a meta-evaporitic origin (e.g., Moine et al. 1981) while variation in spinel chromophore concentrations can affect spinel color, and therefore its value as a gemstone (see discussion below). Metal/Al fractions are not an ideal method of comparison in the case of Lake Harbour Group spinel-bearing rocks since only some trace metals correlate with Al and the metasediments were probably a mixture of carbonate and siliciclastic sediments. Therefore, a 113  qualitative method of estimating the expected trace element composition of the metasediment samples is employed by calculating the expected relative trace element contribution of the various protolith components (estimated above) using trace element averages for sandstone, limestone (Turekian & Wedepohl 1961) and shales/clays (Parker 1967). Concentrations are corrected for the expected mass loss due to devolatilization during metamorphism estimated in Table 4.16 by the sum of the whole rock composition and additional CO2 and H2O estimated to have been lost by devolatilization. The qualitative “expected” trace element concentrations and the actual values relative to these expected concentrations (presented as a fraction) are shown in Tables 4.17A-B. 4.5 Discussion 4.5.1 Origin of spinel occurrences, parageneses, P-T conditions and timing Spinel occurrences, with the exception of Markham Bay localities (and perhaps Waddell Bay), are interpreted to be metasedimentary (sensu stricto) due to their isolation within marble and paragneiss units with no proximal intrusive bodies or signs of large-scale metasomatism. Spinel at Markham Bay is interpreted to be of metasomatic origin. 4.5.1.1 Markham Bay A. Spinel Island Diopside, spinel, phlogopite, calcite and local anorthite form a stable assemblage with highly localized retrograde alteration consisting of tremolite and corundum with subordinate clinochlore, margarite, and clinozoisite/zoisite. Diopsidite is interpreted to have formed by the metasomatic reaction between feldspathic rock and marble; this metasomatic diopsidite band is thickest, best developed, and richest in spinel at the contact with pegmatitic syenogranite. Spinel is rare or absent in diopsidite contiguous to the foliated granitoid rocks. The occurrence of 114  diopsidite between marble and syenogranite suggests two possibilities: the intrusion of syenogranite into marble resulted in contact metasomatism forming diopsidite or, alternatively, that fluid influx at the syenogranite-marble boundary at high metamorphic grades resulted in a bimetasomatic reaction. Spinel genesis from the bimetasomatic reaction between feldspathic rock and dolomite-bearing marble is known from the Grenville Province in Québec (Belley et al. 2016). The locally more-extensive diopsidite development associated with the syenogranite could suggest that contact metamorphism resulted in the most intense metasomatism contiguous to the intrusion, and less so along nearby country rocks, however the presence of marble with no visible diopsidite on the middle western side of the syenogranite contradicts this hypothesis. Therefore, I propose that the diopsidite formed by structurally-controlled fluid influx at syenogranite/granitoid rock and marble boundaries resulting in a bimetasomatic reaction between these rocks. This reaction may have occurred at peak or near peak metamorphic conditions; similar mineral assemblages (e.g., pseudo-section in Fig. 4.20) are stable at these conditions; but formation of the diopsidite during M2 retrograde metamorphism (ca. 720 °C, 6.2 kbar, St-Onge et al. 2007) cannot be ruled-out with the available data. The syenogranite pegmatite may be related to the regionally significant Cumberland Batholith, but a 10 kg sample produced no zircons for age dating and therefore the timing of intrusion could not be determined. Chlorite, margarite, corundum and tremolite formed from the retrograde break-down of spinel and diopside following influx of H2O and CO2. B. Unnamed Island Spinel on the unnamed island occurs in parts of very coarse calcite veins that are free of amphibole. Elsewhere in the outcrop, coarse to pegmatitic amphibole occur within a crosscutting network of calcite veins in calc-silicate rock suggesting significant metasomatism. This rock unit 115  bears great resemblance to the metasomatic calc-silicate unit on the nearby Aliguq Island (Butler 2007). 4.5.1.2 Glencoe Island A.  Main occurrence Spinel occurs in dispersed calc-silicate pods within a band of serpentine marble in a psammite sequence – it is clearly of metasedimentary origin. The pods are assumed to have been a siliceous layer within the marble that was boudinaged and displaced during metamorphism. The presence of forsterite in the marble, inferred from subhedral crystals replaced completely by serpentine, and of orthopyroxene in the psammite, indicates a high grade of metamorphism. Microscopic zircon crystals are common in the calcite-rich portion of one spinel-phlogopite-diopsidite sample. A T-X pseudo-section generated with Perple_X (Fig. 4.20) shows that a diopside-spinel-phlogopite-calcite-(dolomite) assemblage is stable at M1A peak metamorphic conditions. B. Marble beach occurrence Calcite, phlogopite, diopside and spinel form a stable assemblage, which is relatively similar to the peak metamorphic assemblage at the main occurrence. The calcite-rich silicate pods are surrounded by marble (> 30 m), clearly indicating a metasedimentary origin. C. Contact occurrence Spinel and phlogopite-bearing diopsidite occurring between marble and gneiss is interpreted to have formed by contact metasomatism during metamorphism.  D. Marble gulch occurrence The forsterite-spinel assemblage in a layer of this marble outcrop is consistent with granulite facies peak metamorphism. Retrograde break-down of forsterite-calcite to diopside-116  dolomite, and the later alteration of diopside to pargasite subsequently occurred. In the retrograde reactions, equilibrium was not achieved on < 1 cm scale, even though parts of the marble layer contain large (6 mm) euhedral pargasite crystals. The lack of alteration in spinel suggests that it was stable during the retrograde formation of diopside and pargasite. 4.5.1.3 Kimmirut area A. Soper River occurrence Diopside-dolomite pseudomorphic mixtures in marble at Soper River probably formed from the break-down of forsterite and calcite, either because of increasing XCO2 or decreasing metamorphic grade (Fig. 4.21). The stability of a forsterite-pargasite-spinel-bearing assemblage (as observed in the samples, with forsterite inferred from the diopside-dolomite replacements after subhedral mineral grains) in Soper River marble is possible at peak metamorphic conditions and 0.3 ≤ XCO2 ≤ 0.5.  B. Soper Falls occurrence Spinel-bearing humitite could not be modeled with Perple_X due to the absence of humite in the thermodynamic database. Given the significant development of humite and its high volatile content (which are not likely present in such abundance during retrograde metamorphism, e.g., retrograde assemblages described by Belley et al. 2017, St-Onge et al. 2007), it is probable that humite-spinel formed during prograde/peak metamorphism. Fluorine stabilizes humite-group minerals at high temperature (Grützner et al. 2015), and “[t]he stability fields for the individual humite minerals expand to more CO2-rich fluid compositions with increasing fluorine content and decreasing total pressure” (Rice 1981). Retrograde dolomite and pargasite post-date the humite-spinel assemblage. Diopside and talc cannot be accurately placed into the paragenetic sequence. Diopside is weakly altered to tremolite, and spinel to clinochlore 117  indicating a minor amount of retrograde alteration to lower metamorphic grade assemblages. Magnesite appears to be part of the humite-spinel assemblage. C. Soper Lake Camp marble The peak metamorphic stable assemblage is interpreted to consist of forsterite, spinel, phlogopite, and pyrrhotite based on petrography, and the presence of forsterite is consistent with high-temperature granulite facies peak metamorphism. One amphibole crystal occurs in contact with forsterite, and does not appear to consume the forsterite; however, the small sample size of Fo-Amph contacts is insufficient to infer a stable assemblage. D. Soper Lake mica mine The exceptional crystal sizes (i.e., phlogopite to 36 cm across) indicate a significant amount of element transport compared to other LHG metacarbonates. The coarse mineralized layer is poorly exposed due to mining and rock exposure is limited due to its proximity to Soper Lake. However, the lack of intrusive rocks in the vicinity suggests that the occurrence is metasedimentary; it is not known if the extreme coarsening could possibly have been the result of fluid influx from an unknown source. Apatite- and scapolite-bearing pegmatitic recrystallized marble occurs several kilometers to the east. These pods are extremely isolated within the marble sequence, which shows that localized extreme coarsening occurs in this region without a clear indication of intrusive rocks or metasomatic involvement. E. Qila occurrence In the calc-silicate pod at Qila (3E-1), pargasite, scapolite, spinel, phlogopite, and carbonate (dolomite, potentially calcite) form a stable assemblage. Spinel, calcite, phlogopite, and pargasite form a stable assemblage in the carbonate-rich zone below it (sample 3E-2). In the contiguous spinel-bearing marble, forsterite forms a stable assemblage with phlogopite, but 118  spinel was not observed in direct contact with either mineral due to the dispersed nature of non-carbonates. Both diopside and pargasite occur in contact with spinel and as isolated grains. Forsterite locally is rimmed by diopside coronae, some of which have an outer pargasite corona. At peak metamorphic conditions (810 °C, 8.0 kbar, St-Onge et al. 2007), a T-X pseudo-section (Fig. 4.22) indicates that forsterite, spinel, phlogopite, calcite, and dolomite form a stable assemblage at peak metamorphic conditions with XCO2 between ~ 0.07 and 0.27. An increase in XCO2 or decrease in metamorphic grade led to the formation of diopside and pargasite in a stable assemblage with forsterite, spinel, phlogopite, calcite and dolomite without crossing the spinel-out boundary. This new equilibrium assemblage led to the partial replacement of forsterite by diopside and pargasite. F. Trailside occurrence Spinel at Trailside occurs in two rock types, a silicate-rich rock (3F-1) and a carbonate-rich one (3F-2). In the silicate-rich unit, phlogopite, calcite, spinel, and a sericite-replaced unknown form a stable assemblage. The retrograde assemblage consists of muscovite, chlorite, calcite, albite, and corundum. It is hypothesized that sericite alteration completely replaced feldspar, scapolite or nepheline, which are known to occur in Kimmirut-area metacarbonates (Belley et al. 2017). Sericite replacement is significant, and may have resulted in a change in bulk composition, making thermodynamic modelling of this unit unreliable. Both scapolite and K-feldspar occur in trace quantities in sample 3F-2. A T-X pseudosection for a representative sample of the phlogopite-calcite-spinel rock (3F-2; Fig. 4.23) shows that spinel was stable at peak metamorphic conditions. Unidimensional analysis of the predicted modal composition of the rock along a XCO2 path indicates that assemblages are dominantly composed of phlogopite and carbonate, with subordinate spinel and 119  other accessory minerals. At peak metamorphic conditions, scapolite and spinel are only expected to occur coevally at very high XCO2. It is also possible that an accessory silicate was replaced by scapolite during retrograde metamorphism. Corundum formed at the expense of spinel during retrograde metamorphism. 4.7.1.4 Waddell Bay occurrence The Waddell Bay spinel locality was not studied in the field, and thus little information can be extracted from the present samples. Phlogopite, spinel, diopside, and calcite formed coevally based on textural relations. The coarse-grained orange calcite is reminiscent of metasomatic calcite vein occurrences in the Grenville Province (P.M. Belley field observations in the Wakefield, Bryson, and Otter Lake areas, Quebec). 4.7.1.5 Unspecified occurrence on Hall Peninsula Forsterite, spinel, and phlogopite are interpreted to have formed at peak metamorphic conditions followed by pargasite and diopside formation by either an increase in XCO2 or a decrease in metamorphic grade, similar to that observed in Qila marble. 4.5.2 Metasedimentary spinel protoliths 4.5.2.1 Protolith composition of calc-silicate rocks and marbles Metasedimentary (sensu stricto) spinel occurrences in the LHG consist of magnesitic marl (Soper Falls humitite), dolomitic marl (calc-silicate rocks), and dolomitic and dolomite-bearing limestone protoliths (Fig. 4.25; Table 4.16A and B). Interestingly, other significant gemstone occurrences in the LHG are classified as meta-marl: the Beluga sapphire protolith is interpreted to be dolomitic marl (Belley et al. 2017), and the Soper River lapis lazuli an evaporite-bearing marl (Hogarth & Griffin 1978, present study). Since all three gem minerals 120  (spinel, corundum, and the feldspathoid haüyne) occur in silica-undersaturated rocks, high Al/Si (therefore siliciclastic mud but not sand) in marl protoliths would favor gem mineral genesis. Metamorphic reactions between intermixed carbonate and muddy sediments favor Si undersaturation at Al/Si abundances that would not result in corundum or spinel genesis in a metapelite with an identical Al/Si fraction (i.e., in the metapelite with identical Al/Si, silica would be saturated). Nonetheless, the effect of P-T history on the mineral assemblage should not be disregarded, since it can also be an important control on whether spinel or corundum forms at all; this is true of the Beluga sapphire occurrence, where sapphire formation was only made possible by the localized retrograde break-down of nepheline and scapolite. While most spinel-bearing silicate rocks have Al:Si ratios corresponding to mud/clay (see Table 4.16, Fig. 4.24), spinel-bearing rock at the Trailside occurrence is exceptionally rich in Al relative to Si and is similar in Al/Si composition to kaolinite-rich claystone (e.g., claystone sample AR-41 of López et al. 2005, Al/Si ≈ 0.7 g/g). The Trailside spinel-bearing rock, diopsidite, and diopsidite in the Trailside area contain elevated K/Al (0.4-0.8 mol/mol) due to their higher phlogopite content. This elevated K content relative to Al may be due to initial abundances of K-feldspar or illite in the protolith, or from K infiltration (a highly mobile element) with fluid during diagenesis/metamorphism. Trailside K/Al molar fractions are significantly higher than those of other calc-silicate rocks in the LHG (K/Al ≤ 0.3), average clay/shale (0.2, Parker 1967, Turekian & Wedepohl 1961), and sandstone (0.3, Turekian & Wedepohl 1961). López et al. (2005) noted arenite with K/Al = 0.9, likely owing to high K-feldspar content in the sandstone. Another possible origin for potassium enrichment in the spinel-bearing rocks is that it was sourced from the contiguous diopside-rich unit inferred to be metamorphosed sandstone (i.e., the sandstone could have had a relatively high K/Al). 121  The relative proportion of sand and mud (estimated based on Al/Si relative abundance) does not appear to affect the occurrence of spinel so long as Al is sufficiently abundant in the rock; there is considerable overlap in the estimated protolithic composition of spinel-bearing, phlogopite-rich spinel-absent, and phlogopite-poor spinel-absent marbles (Fig. 4.25). Spinel genesis in marbles in the LHG samples is limited by high K and low Mg activities, as discussed below. A lack of Si relative to Al in a K-rich magnesian marble would also result in spinel formation. LHG calc-silicate rock samples (including most spinel-bearing silicate rocks) have roughly the same molar abundance in Ca relative to Mg (Fig. 4.26) and thus the original carbonate fraction in the protolith is interpreted to have been predominantly dolomite (Tables 4.16A and B). Spinel-bearing humitite from Soper Falls is exceptionally Mg-rich and Ca-poor; it is interpreted to have been a magnesite-rich marl layer (an abundance of magnesian clays in the protolith could also explain the high Mg content), however the poor exposure did not allow for sampling of carbonate rocks in contact with the humitite. The possibility that Mg was enriched in the humitite unit and depleted in the contiguous marble should not be disregarded. Marble samples range from dolomitic to calcitic; most consist of dolomite-rich calcite marble. 4.5.2.2 Spinel and meta-evaporite  Garnier et al. (2008) and Giuliani et al. (2018) noted a relationship between selected metacarbonate-hosted gem deposits and evaporites (i.e., lapis lazuli and East Asian ruby deposits). Giuliani et al. (2018) suggested that the presence of meta-evaporites or their indicators (e.g., high salinity fluid inclusions, lazurite/haüyne, marialitic scapolites, tourmalinites, aspidolite, Warren 2016) could be used as exploration criteria for gem deposits in metamorphosed platform carbonate sequences.  Hogarth & Griffin (1978) convincingly showed 122  that the Soper River lapis lazuli is a meta-evaporite citing the following evidence: (1) the lapis lazuli has well-developed layering parallel to the regional foliation, suggesting that it is metasedimentary; (2) the area has a scarcity of intrusive rocks, which does not support a contact metasomatic origin; and (3) the abundances of Na, K, S, Cl, Br, F, and Fe are consistent with evaporite-related sediments. Indeed, the Na and Cl concentrations in Soper River lapis lazuli (this study) are extraordinarily high (i.e., ~6.6 wt. % Na2O and 3170 µg/g Cl) relative to other Lake Harbour Group meta-marls (≤ 2.71 wt. % Na2O and ≤ 1160 µg/g Cl, at Beluga sapphire and Qila, respectively) and a hypothetical shale-limestone mixture of average composition.   Most spinel occurrences on Baffin Island are interpreted to have dolomitic limestone and dolomitic marl protoliths that are consistent with typical non-evaporitic platform sediments transformed by subsequent diagenetic effects (i.e., dolomitization). Evaporitic rocks typically have high Mg contents, relatively low Fe (caused by the widespread occurrence of Mg-rich clays), and evaporitic argillites are characterized by high K, Li, F, and B contents (with the exception of low Li concentrations in Mg-rich evaporites), and low Na contents (with the exception of halite-bearing evaporites; Moine et al. 1981). The abundance of Mg relative to Ca at most spinel occurrences is adequately explained by diagenetic dolomitization and does not imply an evaporitic origin. Compared to expected concentrations for a similar non-evaporitic protolith of “average” composition, metacarbonates are generally poorer in Li than expected (Tables 4.17A and B). Spinel-bearing rocks at Qila, Trailside, Soper Falls, and in Soper Lake mine spinel-bearing marble are richer than expected in B, F, and Cl. Spinel-bearing rock at the main occurrence on Glencoe Island and Trailside area marble are also higher than expected in F and Cl. High F and Cl contents reflect the presence of pargasite, phlogopite, humite, and/or scapolite. While the latter rock samples are enriched in F and Cl, these elements are volatiles 123  (expected to be highly mobile) and may not be representative of the protolith. Moreover, pore fluid salinity can significantly increase at high grades of metamorphism, such as in the amphibolite facies metamorphism of impure marbles and progressive granulite facies metamorphism (Yardley & Graham 2002), which would be expected to lead to increased incorporation of F and Cl relative to OH in phlogopite, pargasite and humite. In addition, a lack of mineral species that can readily incorporate evaporite-sourced volatiles may preclude a rock from retaining its evaporitic geochemical signature during metamorphism.  The composition of humitite at Soper Falls, compared to other LHG rock samples analyzed for the same elements, is much more extreme in concentrations of boron (304 µg/g), fluorine (> 2 wt. %), and magnesium (46.4 wt. % MgO), which we interpret as sufficiently above expected values so as to indicate a probable meta-evaporitic origin consistent with the criteria of Moine et al. (1981).   The presence of evaporites at most spinel occurrences is therefore unlikely, and evaporites are not genetically related to gem spinel: metamorphism of a protolith with the correct proportions of major elements, which occur in typical non-evaporitic carbonate platform sedimentary rocks, is the only criteria. Some geochemical aspects of evaporites noted by Moine et al. (1981) would, in theory, have a negative effect on gem spinel potential (i.e., K enrichment, see Geochemical factors in spinel genesis below) or a positive one (i.e., low Fe content, see Controls on spinel color below). 4.5.3 Geochemical factors in spinel genesis Spinel-bearing calc-silicate rocks are richer in Al relative to Si compared to calc-silicate rocks devoid of spinel (Fig. 4.27A) in spite of having similar relative concentrations of Ca and Mg (see section above). The relative abundances of Al and Si appears to be an important control 124  on whether spinel could form; calc-silicate rocks with low Al/Si are at best silica-saturated (no silica undersaturated phases such as Al oxides, forsterite, haüyne; e.g., sulphide-rich diopsidite at Glencoe Island), and at worst, are silica oversaturated (e.g., quartz-bearing diopsidite near Trailside). Most calc-silicate rocks have K/Al molar fractions of 0.1-0.3 while the Trailside spinel-bearing rock, diopsidite, and diopsidite in the Trailside area contain elevated K/Al (0.4-0.8) due to their higher phlogopite content. Since all calc-silicate rocks studied contain Al > K, potassium activity is not expected to control whether or not spinel occurs. In silicate-rich metacarbonates adequate for thermodynamic modelling (Figs. 4.20, 4.23), spinel forms part of the stable mineral assemblage across the range of possible fluid XCO2 at peak metamorphic conditions. Therefore, the dominant geochemical control on spinel genesis in magnesian calc-silicate rocks under these P-T conditions (granulite facies, 810 °C and 8.0 kbar) appears to be the abundance of Si relative to Al. One exception applies to Qila (described below). In contrast with calc-silicate rocks, spinel-bearing and spinel-absent marbles overlap in Al/Si and Ca/Mg ratios, but differ significantly in K/Al molar ratios. Spinel-bearing marbles are all very poor in potassium, while other marbles contain K/Al ≈ 1 (Fig. 4.27B). Phlogopite is part of stable mineral assemblages with spinel and forsterite, diopside, or pargasite. The proportions of these minerals are expected to obey the following equilibrium reactions: (1) 3 Di + Spl + 2 Dol + 2 K+(aq) + 4 (H2O,F) + CO2 ↔ 2 Phl + 5 Cal (2) 6 Fo + Spl + 7 Cal + 2 K+(aq) + 4 (H2O,F) + 7 CO2 ↔ 2 Phl + 7 Dol (3) 6 Prg + 3 Spl + 9 Dol + 12 K+(aq) + 12 (H2O,F) + 3 CO2 ↔ 12 Phl + 21 Cal In all three reactions, low K activity would favor spinel over phlogopite. The forsterite-spinel marbles contain sufficient calcite for the reaction to proceed, and thus K could be a limiting reactant preventing complete incorporation of Al into phlogopite. Similarly, in some 125  phlogopite-bearing marbles and in Qila calc-silicate rock, dolomite, pargasite, and spinel occur as a stable assemblage, leaving low K activity as a potential limiting factor in the phlogopite-forming reaction. The predominance of Na over K in the pargasites indicates a Na-dominant fluid composition and reinforces the low-K hypothesis. At the contact metamorphic occurrence at Spinel Island, spinel-, phlogopite-, and calcite-bearing diopsidite formed by the metasomatic reaction of a K-Al-Si-rich rock (syenogranite pegmatite, K/Al = 1.2) and phlogopite-bearing calcite marble. The contact metamorphic calc-silicate unit grades into marble, but the contact with unaltered marble has been eroded. However, all of the marble contiguous to the calc-silicate layer has been coarsely recrystallized to calcite with subordinate diopside and phlogopite. Therefore, in diopsidite at Spinel Island, insufficient Mg/dolomite rather than potassium probably prevented complete Al incorporation into phlogopite, which enabled the formation of spinel. Another key factor in spinel genesis in marbles is metamorphic fluid composition:  at peak metamorphic conditions in the LHG, spinel only occurs at relatively low XCO2 (Fig. 4.21-4.22) These controls on the presence or absence of spinel are by no means exhaustive, since mineral assemblages vary in composition (e.g., scapolite at Qila; muscovite replacements after possible K-feldspar at Trailside; humite at Soper Falls), and the equilibrium reactions provided above demonstrate that other factors (such as calcite abundance in a forsterite-spinel assemblage) could be important controls on spinel formation in highly Mg-rich rocks. 4.5.4 Origin of cobalt enrichment at cobalt-blue spinel occurrences 4.5.4.1 The distribution of cobalt enrichment and genetic implications The relatively elevated cobalt concentrations at Qila (9-29 µg/g) and Trailside (9-27 µg/g) vary on outcrop scale, but are isolated to these Co-enriched occurrences: (Qila area) one 126  impure marble sample contains only 3 µg/g Co (although Co/Fe and Co/Al are relatively high, see below) and a pale violet spinel from another nearby marble sample contains Co below the detection limit; (Trailside area) marbles contain ≤ 2 µg/g Co, and calc-silicate rocks 3F-CS2 and 3F-CS3 contain 5 and 8 µg/g Co, respectively. At Trailside, spinel in two contiguous lithologies contains similar Co concentrations (0.06 wt. % CoO), while spinel from the main calc-silicate pod at Qila (0.07 wt. % CoO) is much richer in the element than spinel in the pargasite-calcite rock and marble. There does exist some evidence for up to meter-scale trace element redistribution; the most notable example is the elevated scandium concentration in Trailside diopsidite (4 µg/g) relative to mica-, spinel-, carbonate-rock (1 µg/g), probably caused by the scavenging of Sc by diopside (diopside preferentially incorporates Sc relative to the associated minerals; PM Belley, unpublished distribution coefficients calculated from LA-ICP-MS data). However, while cobalt and nickel concentrations vary considerably by rock type, they are still relatively high in spinel-poor or spinel-free rock like phlogopite-rich marble at Qila (3E-3-B) and diopsidite at Trailside (3F-4). Thus, based on low Co content of local marbles and the Qila area spinel, it is abundantly clear that Co enrichment does not permeate metasedimentary rocks in the immediate area (~ 30 m radius) around the occurrences – although diffusion of trace elements may have occurred on a ≤ 1 m scale. Therefore, the vivid blue color of the spinel is made possible by highly localized Co-enrichment that is attributed to the protolith; i.e., cobalt enrichment occurred prior to peak regional metamorphism, either as a result of Co-rich allochthonous sediment input or authigenic enrichment during sedimentation, diagenesis, or low-grade metamorphism. Chauviré et al. (2015) suggested that, at a cobalt-blue spinel occurrence in Vietnam, cobalt and nickel were mobilized from either the carbonate rocks or nearby amphibolitic rocks by evaporite-derived fluids to the spinel-bearing metacarbonate. This 127  explanation is inconsistent with observations made at Baffin Island cobalt-blue spinel occurrences. 4.5.4.2 Trace metal contents at Co-rich occurrences compared to other LHG rocks and expected protolith compositions Cobalt concentrations in rocks at the Qila and Trailside occurrences (9-29 µg/g) are more elevated than in all analyzed LHG metacarbonates in this study (mostly ≤ 6 µg/g, with the exception of a Trailside area calc-silicate rock, 8 µg/g, and a sulphide-rich diopsidite on Glencoe Island, 16 µg/g). Co/Fe and Ni/Fe ratios show a significantly increased differentiation between Qila and Trailside compared to the other rocks (Fig. 4.28). The majority of LHG metasediments appear to form a trend where higher Fe concentrations are accompanied by progressively higher Co. In contrast, Qila and Trailside rock samples are richer in Co, show high variability in Co concentration, and occur in the low Fe range (Fig. 4.29). Cobalt concentrations at Qila and Trailside appear to increase sharply with slight increases in Fe, but the small sample size and high variability limit the reliability of this observation. Interestingly, the trend in Co/Fe vs Ni/Fe at Qila and Trailside has a similar slope to that of claystones, shales, and sandstones studied by López et al. (2005), but shows that cobalt is very strongly enriched relative to nickel (which itself is in the normal range to enriched) at Qila and Trailside (Fig. 4.28). Cobalt and nickel concentrations show no similar relations with Mn content. The Co-rich rocks at Qila and Trailside are generally above average in Co/Al, but are most distinguished from the other LHG metasediments by Co/Fe, which varies by an order of magnitude (Fig. 4.30). Period IV metal content profiles vary considerably between shale, sandstone, and sedimentary carbonates. Shales and clays are richest in elements V to Zn, with the exception of Mn, which is more abundant in platform carbonates: average shales/clays contain 20 µg/g Co 128  and 95 µg/g Ni, in addition to 3.33 wt. % Fe, 670 µg/g Mn, and concentrations of Cu, Zn, Cr, and V from 57 to 130 µg/g (Parker 1967). Average sandstone is poor in Co and Ni (0.3 and 2 µg/g, respectively), Fe (~ 1 wt. %), Mn (50 µg/g), Cu (5 µg/g), and Zn, Cr, and V (16-35 µg/g; Turekian & Wedepohl 1961). Platform carbonates are poorer than the latter rock types in Co (0.1 µg/g), Fe (~0.4 wt. %), Cu, and Cr; with the same V concentration as an average sandstone, but more Ni and Zn than sandstone (Turekian & Wedepohl 1961). Limestone is also typically richer in Mn than both shales and sandstones. Unlike the previous authors, Graf (1962) noted higher average Co concentrations (4.3 µg/g) and Ni concentrations ranging between 7.5 and 17 µg/g in average sedimentary carbonate compositions from different regions. Even when considering the higher cobalt estimate for carbonates, which probably originate from higher Al content (since trace metals are primarily contained within the aluminous fraction of sediment, Schropp & Windom 1988), the calculated estimates for a completely average equivalent to the protoliths at Qila and Trailside contain 1/5 to ½ of the actual cobalt concentration (with two exceptions, sample 3E-2, which is poorer in Co, and 3F-1, in which spinel and phlogopite are under-represented). The main calc-silicate pod at Qila is twice as rich in Co as expected when using the 4.3 µg/g limestone value, despite sampling bias excluding spinel (the principal Co sink). The most anomalously high Co-Ni concentrations are in rocks with low Al/Si (samples 3E-3-A, 3E-3-B, 3F-4); it is possible that Co and Ni were redistributed to the mud-poor protoliths during diagenesis or low-grade metamorphism. In the case of sample 3F-4, the mud content may have been underestimated due to the lower representation of phlogopite relative to most of the diopside-rich zone. Relative to the protolith trace element concentration estimates (Tables 4.17A and B), the rocks at Qila and Trailside are poor in V, Cr, Fe, Mn, Cu, and Zn. The higher than expected Co (and close to expected Ni) concentrations, and lower Fe and Mn are particularly 129  interesting, since Co and Ni are typically incorporated coevally with either iron or manganese in oxides, hydroxides, or sulphides. It is important to note that LHG metacarbonate rocks are generally much poorer in post Period IV transition metals than expected from the calculated estimate (some exceptions in more sulphide-rich rocks, or high Zn at Glencoe Main). This is perhaps due to differences in the trace element composition of LHG sediments relative to the averages used in the estimate. The actual/estimate trace element values for Qila and Trailside are nonetheless much richer in Co-Ni, and poorer in other Period IV metals relative to other metacarbonates in the LHG. 4.5.4.3 Possible effects of metamorphism In a comparative study of a single metasediment formation, Shaw (1954) found relatively constant concentrations of trace elements (including Li, V, Cr, Co, Ni, Cu, Pb) for low, medium, and high-grade metamorphosed rocks of a specific metapelite unit, with a minor decrease of Ni and Cu, and increase of Li and Pb correlating with K-metasomatism. A decrease in Ni could increase Co/Ni ratios, which are discussed below. However, given the results of Shaw (1954), we proceed under the assumption that trace metal concentrations were not significantly modified by high-grade metamorphism. 4.5.4.4 Possible explanations for the trace metal signature at Qila and Trailside Since expected concentrations based on sediment averages do not adequately explain the Co-Ni enrichment (conservatively twice that expected in a similar protolith of average trace metal composition) together with low V, Cr, Fe, Mn, Cu and Zn abundances, the potential source of Co and Ni must have high Co/Al and very high Co/Fe and Ni/Fe ratios. We explore possible sources of Co enrichment. 130  Price (1972) measured maximum concentrations of approximately 900 µg/g Co and 950 µg/g Ni in syngenetic sedimentary pyrites; these are extreme deviations from the average concentrations (41 µg/g Co, 65 µg/g Ni). Even at the maximum outlier value for sedimentary pyrite Co, the Co/Fe ratio (19 × 10-4 g/g) is smaller than at Qila and Trailside (24 × 10-4  to 51 × 10-4 g/g), and sulphides are very rare at these occurrences (i.e., no traces of original sulphides, although this could change with metamorphism). The average Co/Fe for sedimentary pyrites is 0.9 × 10-4 (Price 1972), considerably lower than all LHG metasediments analyzed in this study (1.7 × 10-4 to 8.2 × 10-4). Moreover, Cu concentrations at Qila and Trailside suggest strong Cu depletion relative to the expected protolith composition; this is contrary to expected Cu concentrations in sediment deposited under euxinic conditions, where Cu would be preserved in sulphides. In Paleoproterozoic black shale-hosted sedimentary pyrite, the trace elements Ni, Cu, and As occur in higher concentrations than Co (Gregory et al. 2015). This is consistent with the interpretation that Co was not sourced from sedimentary pyrite. Exceptionally Co-rich pyrite (Co/Fe = 0.011 g/g) containing a relatively low abundance of As, Pb, Zn, and Cu occurs at Pyrite Hill, Australia in a metamorphosed hydrothermal alteration zone within a granulite facies metasedimentary sequence (albite rock [interpreted to be metamorphosed tuffaceous rock], psammite, metapelite, amphibolite; Plimer 1977).  However, the pyrite’s very high Co/Ni (6.7, attributed to a volcanic exhalative origin by Plimer 1977) and its association with hydrothermal quartz are incompatible with a potentially similar mode of enrichment in cobalt-enriched rocks at Qila and Trailside, where Co enrichment is highly localized, the rocks generally have low silica contents, and much lower Co/Ni (0.6 to 1.5).  In sediments, Co and Ni can be leached from lower parts of the sedimentary sequence in suboxic conditions and become enriched in oxic sediments near the sediment-water interface 131  during early diagenesis (Heggie & Lewis 1984, Gendron et al. 1986, Shaw et al. 1990). These oxic conditions simultaneously result in the loss of Cr and V (while several factors influence the behaviour of Cu; Shaw et al. 1990). Stockdale et al. (2010) experimentally examined the association of Co with Fe and Mn in sediment columns, and suggest that Co is significantly more enriched in authigenic Mn oxides than Fe oxyhydroxides, but that Fe-Co maxima in some experiments were probably caused by sulphide formation by sulphate reduction. Co/Fe and Co/Mn ratios in Stockdale et al.’s experiment remain orders of magnitude smaller than that at Qila and Trailside. Overall, Co-Ni enrichment in an oxic zone of sediments during early diagenesis best explains the unusually high Co-Ni and low Cr and V at Qila and Trailside with two exceptions: (1) Co-Ni would need to be sourced from underlying Co-Ni bearing units, but the surrounding rock is dominantly carbonate that is expected to contain little Co and Ni (i.e., host marbles are Co-Ni poor, present study; average limestone is also Co-Ni poor, Turekian & Wedepohl 1961; while sulphides can occur in limestone/marble, high Co-Ni from sulphides would be expected to be accompanied with higher Fe concentrations); and (2) inconsistency with the relatively low Mn concentration. Furthermore, the much smaller Co/Fe and Ni/Fe values, and the apparent correlation of Co and Fe which cannot be adequately explained by the distribution of Fe-Co-rich phases in different lithologies, would suggest the concentration of Co and Ni in a Fe-bearing phase as opposed to Mn. In oxic conditions, if Co and Ni were concentrating primarily in Fe (oxy)hydroxide, higher whole rock Mn contents would nonetheless be expected; for example, particularly Fe-rich shallow marine ferromanganese concretions from the Black Sea have a Mn/Fe mass fraction of ¼ (Table 79 of Carmichael 1989). If Co and Ni were concentrating in authigenic sulphide, higher Cr, V, and Cu concentrations would be expected under these 132  conditions. Therefore, perhaps Co-Ni- and Mn initially concentrated in the oxic layer of sediment, which was later reduced during diagenesis, leading to the concentration of Co and Ni in Fe-sulphide, their diffusion to contiguous mud-poor units, and the loss of some Mn. There are, however, only traces of sulphide present at Qila and Trailside, and so for this hypothesis to be possible, sulphides must have been destroyed during metamorphism. Concentration of Co in Fe-oxide or siderite is rejected as a possible explanation since they would not result in coeval Mn loss.  Interestingly, meta-marl at the nearby Beluga sapphire occurrence shows the opposite geochemical trend to Qila and Trailside; it is enriched in V and Cr, and significantly impoverished in Co and Ni, which is interpreted to be the result of reducing early diagenetic conditions (Belley et al. 2017). An alternative hypothesis is that Beluga rock was initially enriched in Cr, V, Co, Ni (i.e., deposition reduced euxinic conditions with formation of sedimentary pyrites containing Co and Ni) followed by subsequent dissolution of the sulphides during diagenesis or early metamorphism, resulting in loss of Co and Ni. Cobalt enrichment during diagenesis or metamorphism could have been facilitated by high initial sediment Co content due to the sediment source composition. Ultramafic rocks, and laterites derived from their weathering (e.g., Eliopoulos & Economou-Eliopoulos 2000), have elevated Co/Fe and Co/Mn values (e.g., data of Turekian & Wedepohl 1961), however this is accompanied by very low Co/Ni (0.075, Turekian & Wedepohl 1961). Basaltic rocks, compared to other igneous rock types, are more similar to the average trace metal mass fractions for Qila and Trailside in Co/Fe (5.5 × 10-4 vs 35 × 10-4), Co/Mn (320 × 10-4 vs 715 × 10-4), Ni/Fe (15 × 10-4 vs 46 × 10-4), Ni/Mn (870 × 10-4 vs 940 × 10-4), and Co/Ni (0.37 vs 0.86). “Potential sources for detritus of the Lake Harbour Group […] have yet to be identified,” but Sc/Th ratios in LHG 133  metapelites suggest at least partial derivation from ferromagnesian-rich sources (Thériault et al. 2001).  4.5.5 Controls on spinel color 4.5.5.1 Spinel composition Most spinel in Lake Harbour Group metacarbonates are violet, often with greyish overtones; spinel at Spinel Island and Glencoe Beach is dark blue; spinel at Qila and Trailside is vivid blue; spinel at the Soper Lake mica mine is greyish-purple; and spinel in metasomatite on the unnamed island in Markham Bay is black. The purple color in spinel from the Soper Lake mine is probably due iron in combination with its higher than average Cr2O3 concentration (0.14 wt. %), which produces a red hue in spinel even at low concentrations (0.1 wt. %, Kleišmantas & Daukšytė 2016). The dominant chromophore present in violet and blue spinel is assumed to be iron based on visible absorption spectra mesurements on gem spinels (see section 1.5.2). Vivid blue spinel from Qila and Trailside are colored vivid blue by cobalt (0.03 – 0.07 wt. % CoO) and likely has colour contribution of iron; the darker hue in Trailside spinel may be due to higher FeO contents (~ 5 wt. %) relative to Qila (~ 2.3 – 3.6 wt. %). In LHG rocks, spinel is generally darker in color with increasing iron content: (1) lighter colored spinel contains 1.37-3.55 wt. % FeO; (2) spinel with slightly darker saturation contains 2.65-5.20 wt. % FeO, and were these materials sufficiently transparent and cut into gemstones 1 carat or larger, would appear overly dark, thus lowering their marketability; (3) spinel too dark to produce gemstones, 3.33-7.34 wt. % FeO; and (4) black spinel containing 8.61 wt. % FeO. While lower Fe concentrations (max 6 wt. % FeO, ideally < 4 wt. %) favor attractiveness in spinel (dark spinels have low commercial value), However, overlap between Fe concentrations and the 134  categories of spinel attractiveness in hand sample deserves further study (i.e., quantifying visible light absorption spectra) before narrower conclusions can be made. 4.5.5.2 Relation of spinel Fe content to whole rock composition and mineral assemblage Higher concentrations of iron in spinel result in darker coloration, which controls the mineral’s attractiveness and potential value as a gemstone. Therefore, understanding the relationship between whole rock and spinel Fe concentrations could help in understanding rock types (protoliths or reactive components) that have the potential to produce high quality gem spinel. Iron is preferentially partitioned into spinel relative to the associated silicates (i.e., spinel XFe is higher than that for any associated silicate; see EPMA data) and extremely strongly partitioned into spinel relative to carbonates. Paired whole rock and spinel Fe concentrations generally follow a linear correlation in that most spinel, have approximately 4.3× more Fe by weight % (Fig. 4.31). One exception exists: the spinel at Soper Falls is only about 1.5× richer in Fe. This may be explained by the high abundance of humite, in which Fe substitutes for Mg. While Fe is preferentially incorporated into spinel relative to humite, the distribution coefficient is not so significant as to negate the effects of a very high abundance of humite relative to spinel (~ 20× by volume). It was established above that in LHG samples, only spinel with < 6 wt. % FeO appears sufficiently light colored to represent adequate gem material, or < 4.7 wt. % Fe. In most rocks described above, this would correspond to a Fe concentration of 1.09 wt. % Fe or less. “Expected” protolith composition estimates based on average shale, sandstone, and limestone are above the Fe threshold (~ 2 wt. % Fe; Tables 4.17A and B) for spinel-bearing calc-silicate rocks at Glencoe (3A-SPL-1), Qila (3E-2), and Trailside (3F-2), which suggests that lower-than-average Fe contents could be an important control on spinel gem quality in meta-marl. It should 135  be repeated that mineral assemblage and mineral abundances can have a significant effect on the concentration of Fe in spinel, as shown at Soper Falls. Iron sulphides may also limit Fe availability to spinel, and their abundance could favorably improve the gem-quality of spinel as a result of this preferential incorporation.  While the presence of coeval pyrrhotite or pyrite would improve spinel gem quality by preventing the latter from becoming overly dark, it may also limit or prevent the formation of vivid blue cobalt-enriched spinel, since cobalt is strongly partitioned in pyrrhotite relative to spinel (e.g., see spinel-pyrrhotite pairs from the same rock samples in Tables 4.3 and 4.13). 4.5.6 Exploration criteria The most significant gemstone occurrences in the Lake Harbour Group are all hosted in dolomitic meta-marls of variable composition (Beluga sapphire, Belley et al. 2017; Soper River lapis lazuli meta-evaporite, Hogarth 1971, Hogarth & Griffin 1978; spinel-bearing calc-silicates including cobalt-blue spinel). While dolomite-bearing marbles also contain spinel, the significantly higher spinel concentrations in calc-silicate rocks are more suitable for hard rock gem mining (the relative lack of fluvial environments make Baffin Island an unlikely source of placer-hosted gemstones). Regions with abundant calc-silicate layers intercalated with marble are highly prospective for gem deposits, and these layers could potentially be traced stratigraphically. Local variations in Al/Si and K/Al can lead to spinel genesis in rocks otherwise devoid of the mineral. The favorable geology, widespread abundance of metacarbonates, known gemstone occurrences, and excellent rock exposure make southern Baffin Island one of the most significant prospective areas for colored gem exploration – yet, due to its geographic isolation, one that has yet to be explored in any significant capacity. 136  The occurrence of two unique cobalt-blue spinel occurrences in the same area on Baffin Island is very promising in terms of gem potential. However, the origin of cobalt enrichment in metacarbonates remains poorly understood, making targeted exploration difficult.  4.6 Conclusions 4.6.1 Origin of spinel occurrences Spinel occurrences in the Lake Harbour Group mostly consist of metasedimentary (sensu stricto) deposits, with the exception of the two occurrences in Markham Bay, which are metasomatic. The origin of the Waddell Bay spinel occurrence cannot be ascertained but may be metasomatic. Spinel-bearing calc-silicate rock at Spinel Island in Markham Bay formed from the reaction of syenogranite pegmatite with dolomite-bearing marble. All spinels occur in marble and calc-silicate/silicate-rich metacarbonate rocks. Minerals occurring with spinel as part of a stable assemblage include calcite, dolomite, phlogopite, pargasite, diopside, humite, forsterite, scapolite, anorthite, and pyrrhotite. Spinel is interpreted to have formed during peak granulite facies metamorphism. Spinel at Trailside and Spinel Island is locally replaced by retrograde corundum, chlorite, and other minerals. 4.6.2 Metasedimentary protoliths Spinel-bearing metacarbonate rocks are interpreted to have the following protoliths: (1) impure dolomite-bearing and dolomitic limestone; (2) dolomitic marl; and (3) evaporitic magnesitic marl (Soper Falls humitite). The protoliths are relatively similar to the Beluga sapphire occurrence (dolomitic marl with lower Al/Si than in the spinel-bearing rocks) and the Soper River lapis lazuli rock (evaporitic marl). Spinel-bearing calc-silicate rocks have Al/Si relative abundances within range of sandy shale, shale, claystone, and kaolinite-rich claystone. 137  There is no compelling evidence for the presence of evaporites except at Soper Falls, where the humitite is particularly enriched in F, B, and Mg. 4.6.3 Geochemical factors in spinel genesis Spinel genesis in LHG metacarbonates is favored by: (1) the low abundance of silica relative to alumina, the primary control on whether spinel forms in most calc-silicate rocks; (2) low potassium activity limiting the formation of phlogopite and thus leaving Al available for spinel formation; (3) low XCO2 in marbles; and (4) insufficient quantities of Mg or dolomite reactant in diopsidite limiting Al incorporation into phlogopite to form spinel. 4.6.4 Origin of cobalt enrichment at Qila and Trailside The spatial distribution of Co-Ni enrichment at Qila and Trailside and the surrounding (~ 30 m) areas is indicative of highly localized enrichment, with possibly only small-scale (≤ 1 m) diffusion during peak metamorphism. This suggests that trace metal concentrations in cobalt-blue spinel bearing metacarbonates are a result of protolith composition and not due to the infiltration of metal-rich metamorphic fluids. Cobalt and Ni are therefore expected to have been enriched in the original sediment, or during diagenesis or low-grade metamorphism. Concentrations of Co (especially) and Ni are anomalously high (conservatively twice the expected cobalt concentration in a comparable metasediment), while concentrations of Fe, Mn, V, Cr, and Cu are much lower than expected; a chemical signature that, with the exception of low Fe and Mn, could be caused by element enrichment/depletion in the oxic layer of sediment during early diagenesis. I propose the possibility that later sulfide formation within the rock during low-grade metamorphism could have retained Fe, Co, and Ni and resulted in some loss of Mn (which would have been initially enriched together with Co and Ni in oxic conditions). 138  Ultimately, such explanations are speculative since the protoliths have undergone diagenetic, structural, and metamorphic transformations. 4.6.5 Controls on spinel color Vivid blue spinel at Qila and Trailside are enriched in cobalt, containing 0.03-0.07 wt. % CoO. Spinels from other occurrences are blue and violet, and Fe is the probable chromophore. Lighter-colored spinels contain 1.37-3.55 wt. % FeO. Slightly darker spinels, which could only produce small gemstones (stones over a carat would likely be too dark for use as a gem), contain 2.65-5.20 wt. % FeO. Blue and violet spinels that are too dark for use as a gemstone contain 3.33-7.34 wt. % FeO. It is estimated that spinel should have concentrations below 6 wt. % FeO to have a chance of being suitable for use as a commercial gemstone. Spinel Fe concentrations are generally 4.3× higher than in the host rock, with the exception of the very silicate-rich humitite rock at Soper Falls (1.5×). Considering the estimated 6 wt. % FeO cut-off for gem spinel, dolomitic marl protoliths would require < 1.09 wt. % Fe to have potential for gem-quality spinel. This value is lower than that expected for similar protoliths of “average” composition (~ 2 wt. % Fe). Pyrrhotite strongly partitions Fe and Co relative to spinel, therefore an abundance of sulfide is expected to improve the gem quality/attractiveness of spinel by decreasing the amount of Fe incorporated into spinel (thus preventing overly dark colors), but would also prevent the formation of vivid blue, Co-enriched spinel in rocks that have suitable cobalt concentrations. 4.6.6 Exploration criteria Marbles containing abundant metamorphosed dolomitic marl (magnesian calc-silicate) layers offer the best potential for gemstone discoveries in southern Baffin Island. Small, sometimes localized variations in whole rock Al/Si and K/Al could result in the occurrence of 139  spinel. Other significant gemstone occurrences on Baffin Island (sapphire and lapis lazuli) are hosted in relatively similar rock types. Meta-marl-rich units may be traceable stratigraphically. 4.7 Tables Table 4.1: Spinel occurrences studied Region Locality Location # Markham Bay “Spinel” island 1A Unnamed island 1B Glencoe Island Main occurrence 2A Marble beach occurrence 2B Contact occurrence 2C Marble gulch occurrence 2D Kimmirut area Soper River occurrence 3A Soper Falls occurrence 3B Soper Lake Camp marble 3C Soper Lake mica mine 3D Qila occurrence 3E Trailside occurrence 3F Hall Peninsula Waddell Bay occurrence 4 Unspecified marble occurrence 5     140  Table 4.2A: South to North section of lithologies surrounding the west extremity of the diopsidite band, Spinel Island, Markham Bay. Thickness (m) Rock Description > 10 Marble (1A-M) Medium-grained, dolomite-bearing calcite marble with subordinate phlogopite and forsterite. Contains sparsely distributed layers (< 10 cm) containing very coarse-grained clots of phlogopite. A representative sample of pale grey marble contains 5% phlogopite, subordinate forsterite, and trace pyrrhotite. 1.5-2.5 Diopsidite Medium- to coarse-grained, locally pegmatitic greenish gray diopsidite with subordinate phlogopite, locally common calcite, and rare spinel. Calcite occurs in randomly oriented veins up to 15 cm thick and 2 meters long, and pods up to 30 cm in maximum dimension. 1 White Kfs rock Discontinuous mass of white K-feldspar rock containing a few small pods and veinlets of diopsidite. 0.4 Diopsidite Diopsidite, in one area containing a 30 cm-thick pod of marble. 1 C. White Kfs rock Coarse-grained white K-feldspar with trace apatite and allanite. 1.3 Pyroxene-bearing feldspar rock Fine-grained, foliated rock composed, by volume, of 35 % albite, 35 % K-feldspar, 16 % clinopyroxene (XMg ≈ 0.5), 12 % orthopyroxene (XMg ≈ 0.4), 1.5 % ilmenite, 0.5 % pyrrhotite, trace apatite and zircon. > 5 Lineated granitoid rock Pale beige rock composed primarily of fine-grained feldspar (albite, K-feldspar), and coarser-grained quartz as masses elongated along the lineation. Some parts of the rock contain black calcic amphibole porphyroblasts averaging 1 cm (XMg ≈ 0.4), or local enrichments in fine-grained ilmenite.  Trace apatite, biotite, and rare monazite.  Table 4.2B: South to North section of lithologies surrounding the spinel-bearing zone of the diopsidite band, Spinel Island, Markham Bay. Thickness (m) Rock Description > 6 Marble Medium- to coarse-grained phlogopite, and phlogopite-diopside marble. ≥ 0.1 Recrystallized marble Very coarse to pegmatitic calcite containing up to 15 % euhedral greyish-green diopside prisms and phlogopite crystals. 3-4 cm crystals common. Contact with marble eroded. 1.5-2.5 Diopsidite (1A-SPL) Medium-grained to pegmatitic diopside with phlogopite, calcite, local spinel, and anorthite. See text. > 10 Syenogranite pegmatite (1A-SG) Pegmatitic microcline (with some perthite) with subordinate quartz, commonly as graphic intergrowth, and plagioclase. Trace titanite, rutile, allanite, and scapolite. Titanite contains 0.5 wt. % Nb (EDS).  141  Table 4.3A: Average composition of spinel from Markham Bay, Waddell Bay, and Hall Peninsula. Normalized to 32 oxygen atoms per formula unit. Locality Spinel Island Unnamed island Waddell Bay Hall Pen.  Lithology Spl diopsidite Spl-bearing metasomatite Cal-Di-Phl-Spl rock Spl-bearing marble Color Dark blue  Black  Blue-violet Dark grey-violet Sample 1A-SPL  1B  4  5  n 5 σ 4 σ 3 σ 4 σ TiO2 (wt.%) 0.01 0 0.01 0.01 < 0.01 0 0.01 0.01 ZnO 0.32 0.02 0.29 0.01 0.36 0.02 0.07 0.01 Al2O3 70.20 0.16 69.59 0.20 71.47 0.08 70.96 0.33 V2O3 < 0.02  < 0.02  < 0.02 0 0.04 0.01 Cr2O3 < 0.03  0.06 0.01 < 0.03  0.07 0.04 FeO 6.9 0.07 8.61 0.01 3.90 0.02 2.87 0.15 CoO < 0.03  < 0.03  < 0.03  < 0.03  NiO < 0.03  < 0.03  < 0.03  < 0.03  MnO 0.08 0.01 0.08 0.01 0.04 0.01 0.03 0.02 MgO 22.89 0.14 22.17 0.13 24.66 0.09 25.65 0.12 TOTAL 100.40  100.81  100.43  99.70  Ti (apfu)         Zn 0.046  0.042  0.051  0.010  Al 16.114  16.040  16.172  16.093  V       0.006  Cr   0.009    0.011  Fe 1.124  1.408  0.626  0.462  Co         Ni         Mn 0.013  0.013  0.007  0.005  Mg 6.646  6.463  7.058  7.358     142  Table 4.3B: Average composition of spinel from Glencoe Island. Normalized to 32 oxygen atoms per formula unit. Locality Glencoe main Glencoe main Glencoe beach Glencoe contact Glencoe gulch Lithology Spl-bearing pod Spl-bearing pod Spl-bearing marble Spl diopsidite Fo-Prg-Spl marble Color Violet  Dark grey-violet Dark blue  Dark grey-violet Violet  Sample 2A-SPL-1  2A-SPL-3  2B  2C  2D  n 4 σ 3 σ 4 σ 3 σ 4 σ TiO2 (wt.%) < 0.01  < 0.01  0.01 0.02 < 0.01 0 0.01 0.01 ZnO 2.32 0.04 4.25 0.07 0.13 0.03 0.41 0.01 < 0.03  Al2O3 69.43 0.20 69.04 0.11 70.31 0.97 70.04 0.09 71.62 0.20 V2O3 < 0.02  0.04 0.02 < 0.02  < 0.02  0.03 0.02 Cr2O3 0.03 0.01 < 0.03  < 0.03  < 0.03  < 0.03  FeO 4.88 0.09 4.23 0.05 7.34 0.08 7.27 0.05 1.72 0.05 CoO < 0.03  < 0.03  < 0.03  < 0.03  < 0.03  NiO < 0.03  < 0.03  < 0.03  < 0.03  < 0.03  MnO 0.06 0.01 0.04 0.01 0.09 0.02 0.10 0.01 0.04 0.01 MgO 22.72 0.13 22.12 0.12 22.67 0.27 22.26 0.02 26.65 0.14 TOTAL 99.44  99.72  100.55  100.08  100.07  Ti (apfu)           Zn 0.337  0.621  0.019  0.059    Al 16.115  16.094  16.129  16.159  16.092  V   0.006      0.005  Cr 0.005          Fe 0.804  0.700  1.195  1.190  0.274  Co           Ni           Mn 0.010  0.007  0.015  0.017  0.006  Mg 6.670  6.522  6.578  6.496  7.574    143  Table 4.3C: Average composition of spinel from Soper Lake and Soper River area, near Kimmirut. Normalized to 32 oxygen atoms per formula unit. Locality Soper River Soper Falls Soper Lk camp Soper Lk mine Soper Lk mine Lithology Spl-bearing marble Humitite  Spl-bearing marble Fo-Carb-Phl-Spl rock Fo-Spl marble Color Violet  Greyish violet Violet  Dark greyish purple Dark violet Sample 3A-2  3B  3C  3D-1  3D-M1  n 5 σ 5 σ 4 σ 4 σ 3 σ TiO2 (wt.%) 0.02 0.01 0.01 0.01 0.02 0.01 0.01 0.01 < 0.01  ZnO 0.05 0.02 < 0.03  0.04 0.02 0.12 0.01 0.18 0.01 Al2O3 70.76 0.18 71.79 0.15 72.01 0.18 71.36 0.25 71.08 0.14 V2O3 < 0.02  0.04 0 0.04 0.02 0.04 0.02 0.03 0.01 Cr2O3 0.05 0.02 < 0.03  0.07 0.02 0.14 0.02 0.05 0.01 FeO 1.74 0.11 2.04 0.03 1.37 0.02 2.65 0.05 3.33 0.07 CoO < 0.03  < 0.03  < 0.03  < 0.03  < 0.03  NiO < 0.03  < 0.03  < 0.03  < 0.03  < 0.03  MnO < 0.03  0.04 0.02 0.03 0.01 0.05 0.01 0.05 0.01 MgO 26.69 0.07 26.38 0.09 26.62 0.10 26.06 0.06 25.62 0.05 TOTAL 99.31  100.30  100.20  100.43  100.34  Ti (apfu) 0.003    0.003      Zn 0.007    0.006  0.017  0.025  Al 16.031  16.113  16.132  16.061  16.061  V   0.006  0.006  0.006  0.005  Cr 0.008    0.011  0.021  0.008  Fe 0.280  0.325  0.218  0.423  0.534  Co           Ni           Mn   0.006  0.005  0.008  0.008  Mg 7.649  7.489  7.543  7.419  7.323     144  Table 4.3D: Average composition of spinel from Qila and Trailside, Kimmirut area. Normalized to 32 oxygen atoms per formula unit. Locality Qila  Qila  Qila  Qila (area) Trailside  Trailside  Lithology Spl-bearing calc-silicate Prg-Cal rock Spl-bearing marble Dolomitic marble Spl-bearing silicate rock Cal-Phl-Spl rock Color Cobalt-blue Sky blue  Cobalt-blue Pale violet Cobalt-blue Cobalt-blue Sample 3E-1  3E-2  3E-3-A  3E-M1  3F-1  3F-2  n 5 σ 4 σ 3 σ 4 σ 4 σ 4 σ TiO2 (wt.%) < 0.01 0 < 0.01  < 0.01  0.02 0 < 0.01 0 < 0.01 0 ZnO 0.16 0.01 0.05 0 0.05 0.02 0.09 0.01 0.05 0.02 0.13 0.02 Al2O3 71.39 0.26 71.75 0.04 72.42 0.23 71.72 0.23 70.84 0.27 70.56 0.25 V2O3 < 0.02  < 0.02 0 < 0.02  0.04 0.02 < 0.02  < 0.02  Cr2O3 < 0.03  < 0.03  0.04 0.01 < 0.03  < 0.03  < 0.03  FeO 3.55 0.11 2.25 0.03 2.45 0.07 1.64 0.06 5.20 0.09 5.12 0.02 CoO 0.07 0.01 0.03 0.02 0.03 0.01 < 0.03  0.06 0.01 0.06 0.01 NiO 0.04 0.01 < 0.03  0.03 0.02 < 0.03  < 0.03  < 0.03  MnO 0.05 0.02 0.04 0.01 0.03 0.01 0.03 0.02 0.06 0.02 0.06 0.01 MgO 24.78 0.17 26.62 0.21 25.57 0.03 26.82 0.04 23.62 0.08 23.65 0.19 TOTAL 100.04  100.74  100.62  100.36  99.83  99.58  Ti (apfu)       0.003      Zn 0.023  0.007  0.007  0.013  0.007  0.019  Al 16.184  16.059  16.223  16.070  16.199  16.180  V       0.006      Cr     0.006        Fe 0.571  0.357  0.389  0.261  0.844  0.833  Co 0.011  0.005  0.005    0.009  0.009  Ni 0.006    0.005        Mn 0.008  0.006  0.005  0.005  0.010  0.010  Mg 7.106  7.536  7.246  7.601  6.832  6.860   145   Table 4.4: Average composition of forsterite associated with spinel, southern Baffin Island. Normalized to 4 oxygen atoms per formula unit. Locality Soper Lk camp Soper Lk mine Soper Lk mine Qila  Qila  Hall Pen.  Lithology Spl-bearing marble Fo-Carb-Phl-Spl rock Fo-Spl marble Spl-bearing marble Spl-bearing marble Spl-bearing marble Sample 3C  3D-1  3D-M1  3E-3-A (green Fo) 3E-3-A (grey Fo) 5  n 3 σ 4 σ 3 σ 3 σ 3 σ 4 σ SiO2 (wt.%) 42.61 0.10 42.21 0.41 41.38 0.28 42.52 0.44 42.18 0.36 43.33 0.19 TiO2 0.01 0.01 0.01 0.01 < 0.01  < 0.01  0.01 0.01 < 0.01  FeO 1.29 0.05 2.55 0.02 3.10 0.08 2.39 0.04 2.28 0.02 2.76 0.06 CoO < 0.03  < 0.03  < 0.03  0.03 0.01 < 0.03  < 0.03  MnO 0.05 0.02 0.09 0.02 0.09 0.01 0.05 0.01 0.07 0.01 0.07 0.01 MgO 56.96 0.05 55.72 0.22 54.58 0.34 55.66 0.42 55.40 0.48 56.07 0.23 CaO 0.02 0.01 0.02 0.01 0.01 0.01 0.02 0.01 0.01 0.01 0.01 0.01 TOTAL 100.97  100.61  99.19  100.70  99.99  102.26  Si (apfu) 0.995  0.995  0.992  1  0.999  1.004  Ti 0  0  0  0  0  0  Fe 0.025  0.050  0.062  0.047  0.045  0.053  Co 0  0  0  0.001  0  0  Mn 0.001  0.002  0.002  0.001  0.001  0.001  Mg 1.983  1.958  1.951  1.951  1.955  1.937  Ca 0.001  0.001  0  0.001  0  0  CATSUM 3.005  3.006  3.007  3.001  3.000  2.995  Fo 0.99  0.98  0.97  0.98  0.98  0.97  Below detection limit: Na2O, V2O3 and Al2O3 (0.02 wt. %); Cr2O3, ZnO, NiO (0.03 wt. %).   146  Table 4.5A: Composition of diopside associated with spinel, Markham Bay and Glencoe Island. Normalized to 6 oxygen atoms per formula unit. Locality Spinel Island  Unnamed Island  Glencoe main  Glencoe beach  Glencoe contact  Lithology Spl diopsidite  Spl-bearing metasomatite  Spl-bearing pod  Spl-bearing marble  Spl diopsidite  Sample 1A-SPL  1B  2A-SPL-3  2B  2C  n 4 σ 4 σ 3 σ 3 σ 3 σ SiO2 (wt.%) 51.02 1.36 51.36 0.22 53.24 0.95 53.25 0.43 51.69 0.32 TiO2 0.17 0.05 0.33 0.02 0.14 0.01 0.31 0.01 0.14 0.02 Al2O3 3.41 0.67 4.62 0.23 3.63 0.21 2.88 0.03 5.07 0.09 V2O3 < 0.02  < 0.02  0.05 0.02 < 0.02  < 0.02  FeO 1.20 0.02 1.63 0.03 0.87 0.03 1.15 0.03 1.21 0.04 MnO 0.03 0.02 0.05 0.02 < 0.02  0.06 0.02 0.05 0.02 MgO 16.35 0.33 15.80 0.20 16.61 0.27 17.01 0.26 15.66 0.15 CaO 25.81 0.08 25.46 0.05 25.54 0.01 25.22 0.07 25.54 0.09 Na2O 0.06 0.01 0.10 0.01 0.02 0 0.06 0.01 0.06 0.02 TOTAL 98.07  99.38  100.13  99.97  99.44  Si (apfu) 1.896  1.882  1.924  1.930  1.887  Ti 0.005  0.009  0.004  0.008  0.004  Al 0.149  0.200  0.155  0.123  0.218  V 0  0  0.001  0  0  Fe 0.037  0.050  0.026  0.035  0.037  Mn 0.001  0.002  0  0.002  0.002  Mg 0.906  0.863  0.895  0.919  0.852  Ca 1.028  1.000  0.989  0.980  0.999  Na 0.004  0.007  0.001  0.004  0.004  CATSUM 4.026  4.013  3.995  4.001  4.003  Below detection limit: ZnO (0.04 wt. %); Cr2O3, CoO, and NiO (0.03 wt. %).    147  Table 4.5B: Average composition of diopside associated with spinel, Kimmirut area, Waddell Bay, and Hall Peninsula. Normalized to 6 oxygen atoms per formula unit. Locality Soper River  Soper River  Soper Falls  Qila Waddell Bay  Hall Pen. Lithology Spl-bearing marble  Spl-bearing marble  Humitite  Spl-bearing calc-silicate Cal-Di-Phl-Spl rock  Spl-bearing marble Sample 3A-1  3A-1 (pseudomorph)  3B  3E-1 4  5 n 3 σ 2 σ 3 σ 1 4 σ 1 SiO2 (wt.%) 55.52 0.32 55.6 0.23 56.01 0.08 52.71 52.83 0.57 53.85 TiO2 0.06 0.01 < 0.01  0.05 0.01 0.06 0.09 0.01 1.03 Al2O3 0.63 0.04 0.07 0.02 0.62 0.02 3.71 5.53 0.43 2.75 V2O3 < 0.02  < 0.02  < 0.02  < 0.02 < 0.02  < 0.02 FeO 0.22 0.02 0.21 0.01 0.28 0.03 0.41 0.96 0.04 0.43 MnO 0.03 0 0.03 0.01 0.03 0.02 0.03 0.04 0.01 0 MgO 18.11 0.06 18.33 0.24 18.46 0.06 16.64 15.79 0.10 17.34 CaO 25.57 0.05 25.7 0.11 25.71 0.03 24.75 24.43 0.05 25.52 Na2O 0.13 0.01 0.03 0.01 < 0.02  0.37 0.61 0.02 0.05 TOTAL 100.31  100.01  101.2  98.71 100.30  100.99 Si (apfu) 1.995  2.005  1.994  1.927 1.901  1.927 Ti 0.002  0  0.001  0.002 0.002  0.028 Al 0.027  0.003  0.026  0.160 0.235  0.116 V 0  0  0  0 0  0 Fe 0.007  0.006  0.008  0.013 0.029  0.013 Mn 0.001  0.001  0.001  0.001 0.001  0 Mg 0.970  0.985  0.980  0.907 0.847  0.925 Ca 0.984  0.993  0.981  0.969 0.942  0.978 Na 0.009  0.002  0  0.026 0.043  0.003 CATSUM 3.995  3.995  3.991  4.005 4.000  3.990 Below detection limit: ZnO (0.04 wt. %); Cr2O3, CoO, and NiO (0.03 wt. %).148  Table 4.6A: Average composition of amphibole from Markham Bay, Glencoe Island, and Hall Peninsula. Results calculated using the spreadsheet of Locock (2014) for 24 anions. Locality Spinel Island Glencoe gulch Glencoe gulch Hall  Lithology Altered Spl diopsidite Fo-Prg-Spl marble Fo-Prg-Spl marble Spl-bearing marble Sample 1A-SPL-Alt 2D 2D  5  n 4 σ 1 4 σ 4 σ Species Tremolite Pargasite Pargasite  Pargasite  SiO2 (wt.%) 56.48 0.87 42.65 44.36 0.18 44.75 0.65 TiO2 0.05 0.05 2.64 1.69 0.09 1.81 0.37 Al2O3 2.22 1.15 15.29 14.21 0.29 14.26 1.25 V2O3 < 0.02  0.06 0.03 0.01 0.02 0.01 FeO 2.76 0.33      Fe2O3   0.43 0.60 0.02 0.89 0.01 MnO 0.05 0.01 < 0.03 < 0.03  < 0.03  MgO 21.74 0.30 19.21 19.12 0.11 19.23 0.21 CaO 13.52 0.08 12.71 13.56 0.04 13.08 0.09 BaO < 0.06  < 0.06 0.21 0.04 < 0.06  Na2O 0.16 0.07 2.92 1.37 0.07 2.20 0.29 K2O 0.09 0.09 0.44 1.05 0.03 0.48 0.09 Cl 0.01 0.01 0.14 0.16 0.01 0.04 0.01 F < 0.12  < 0.12 < 0.12  < 0.12  H2O (calc) 2.17  1.49 1.71  1.72  O=F,Cl 0  -0.03 -0.04  -0.01  TOTAL 99.25  97.95 98.03  98.47        T       Si (apfu) 7.808  6.122 6.340  6.345  Al 0.192  1.878 1.660  1.655  C       Ti 0.005  0.285 0.182  0.193  Al 0.169  0.708 0.733  0.728  V   0.007 0.003  0.002  Fe3+   0.047 0.065  0.095  Mn2+ 0.006       Fe2+ 0.319       Mg 4.480  3.953 4.017  3.982  B       Fe2+        Mg   0.157 0.056  0.083  Ca 2.000  1.843 1.955  1.917  Ba    0.012    Na 0.000  0.000 0.000  0.000  A         Ca 0.003  0.112 0.121  0.070  Na 0.043  0.813 0.380  0.605  K 0.016  0.081 0.191  0.087  W        OH 1.987  1.395 1.597  1.604  F        Cl 0.002  0.034 0.039  0.010  O 0.010  0.571 0.364  0.387  Below detection limit (0.03 wt. %): ZnO, Cr2O3, CoO, NiO.   149  Table 4.6B: Average composition of amphibole from the Kimmirut area. Results calculated using the spreadsheet of Locock (2014) for 24 anions. Locality Soper River Soper River Soper Falls Soper Falls Lithology Spl-bearing marble Spl-bearing marble Humitite Humitite Sample 3A-2  3A-1  3B  3B  n 7 σ 4 σ 3 σ 3 σ Species Pargasite Pargasite Tremolite Pargasite SiO2 (wt.%) 46.03 1.07 45.10 0.93 56.54 0.53 46.39 0.50 TiO2 0.23 0.09 0.58 0.44 0.10 0.03 0.14 0.04 Al2O3 14.03 1.78 14.00 1.05 3.98 0.45 14.20 0.95 V2O3 0.02 0.01 0.02 0.01 0.02 0.01 0.06 0.02 FeO 0.23 0.01   0.34 0.02 0.41 0 Fe2O3 0.25 0.01 0.58 0.05 0  0  MnO < 0.03  < 0.03  0.03 0.02 < 0.03  MgO 20.68 0.52 19.89 0.27 22.94 0.23 20.34 0.19 CaO 12.42 0.31 13.10 0.30 13.92 0.06 13.91 0.14 BaO < 0.06  < 0.06  < 0.06  0.07 0.02 Na2O 3.00 0.25 2.61 0.23 0.06 0.03 3.61 0.26 K2O 0.49 0.12 0.48 0.22 0.02 0.01 0.18 0.02 Cl 0.08 0.04 0.18 0.06 0.01 0.01 0.06 0.01 F < 0.12  < 0.12  < 0.12  0.42 0.14 H2O (calc) 2.09  1.97  2.18  1.91  O=F,Cl -0.02  -0.04  0  -0.19  TOTAL 99.53  98.47  100.14  101.51          T         Si (apfu) 6.421  6.380  7.663  6.381  Al 1.579  1.620  0.337  1.619  C          Ti 0.024  0.062  0.010  0.014  Al 0.727  0.714  0.298  0.684  V 0.002  0.002  0.002    Fe3+ 0.026  0.062      Mn2+     0.003    Fe2+     0.039  0.047  Mg 4.220  4.160  4.635  4.171  B          Fe2+ 0.026        Mg 0.080  0.034      Ca 1.856  1.966  2.000  2.004  Ba       0.004  Na 0.037      0.000  A          Ca   0.020  0.021  0.046  Na 0.774  0.716  0.016  0.963  K 0.087  0.087  0.003  0.032  W          OH 1.933  1.833  1.977  1.774  F       0.183  Cl 0.019  0.043  0.002  0.014  O 0.048  0.124  0.020  0.029  Below detection limit (0.03 wt. %): ZnO, Cr2O3, CoO, NiO.  150  Table 4.6C: Average composition of amphibole from the Kimmirut area. Results calculated using the spreadsheet of Locock (2014) for 24 anions. Locality Soper Lk camp Qila Qila Lithology Spl-bearing marble Spl-bearing calc-silicate Prg-Cal rock Sample 3C  3E-1  3E-2  n 3 σ 6 σ 4 σ Species Pargasite Pargasite Pargasite SiO2 (wt.%) 42.50 0.20 41.97 0.38 42.5 0.35 TiO2 2.64 0.05 0.20 0.02 0.20 0.02 Al2O3 15.29 0.28 18.41 0.42 16.47 0.03 V2O3 0.06 0.02 < 0.02  < 0.02  FeO   0.82 0.02   Fe2O3 0.43 0.02   0.70 0.01 MnO < 0.03  0.03 0.02 < 0.03  MgO 19.21 0.16 17.79 0.17 19.14 0.07 CaO 12.71 0.07 12.90 0.04 12.93 0.06 BaO < 0.06  0.12 0.04 0.07 0.02 Na2O 2.92 0.05 2.04 0.13 2.74 0.03 K2O 0.44 0.03 2.23 0.26 2.02 0.03 Cl 0.14 0.01 0.15 0.01 0.14 0.01 F < 0.12  0.20 0.07 < 0.12  H2O (calc) 1.49  1.95  2.05  O=F,Cl -0.03  -0.12  -0.03  TOTAL 97.80  98.69  98.93         T       Si (apfu) 6.112  5.991  6.056  Al 1.888  2.009  1.944  C        Ti 0.286  0.021  0.021  Al 0.703  1.088  0.822  V 0.007      Fe3+ 0.047    0.075  Mn2+   0.004    Fe2+   0.098    Mg 3.958  3.785  4.066  B        Fe2+       Mg 0.160      Ca 1.840  1.973  1.974  Ba   0.007  0.004  Na 0.000  0.021  0.022  A        Ca 0.119      Na 0.814  0.544  0.735  K 0.081  0.406  0.367  W        OH 1.394  1.830  1.923  F   0.090    Cl 0.034  0.036  0.034  O 0.572  0.043  0.043  Below detection limit (0.03 wt. %): ZnO, Cr2O3, CoO, NiO. 151  Table 4.7A: Average composition of phlogopite from Markham Bay, Waddell Bay, and Hall Peninsula. Normalized on the basis of 12 anions per formula unit. Locality Spinel Island Unnamed Island Waddell Bay Hall Pen. Hall Pen.   Lithology Spl diopsidite Spl-bearing metasomatite Cal-Di-Phl-Spl rock Spl-bearing marble Spl-bearing marble Sample 1A-SPL 1B 4 5 5 n 3  σ 4 σ 3  σ 2 σ 5  σ SiO2 (wt. %) 37.81  0.12  37.39  0.25  39.08  0.04  36.27  1.58  37.59  0.79  TiO2 0.28  0.01  0.49  0.01  0.35  0.01  1.60  0.14  1.22  0.13  Al2O3 17.16  0.32  17.03  0.18 17.91 0.02 17.49 0.86  17.19 0.28  V2O3 <0.02   <0.02  <0.02  <0.02  <0.02  FeO 1.91 0.08 2.28 0.06 1.12 0.05 0.91 0.02 0.86 0.02 MgO  24.50 0.21 24.21 0.16 24.95 0.06 24.38 0.64 24.67 0.14 BaO 0.85 0.10 0.99 0.06 0.11 0.05 6.64 1.53 4.30 1.07 Na2O 0.20 0.01 0.54 0.01 0.88 0.06 0.47 0.01 0.56 0.03 K2O 10.12 0.04 9.86 0.03 10.17 0.06 7.00 0.51 7.67 0.34 Cl 0.16 0.01 0.16 0.02 0.03 0.01 0.17 0.02 0.11 0.02 F 1.22 0.18 0.67 0.12 0.42 0.05 <0.12  <0.12  H2Ocalc 3.51   3.75  4.04  4.06  4.12  O=F,Cl  -0.55   -0.32  -0.19  -0.04  -0.02   TOTAL  97.17  97.05  98.87  98.95   98.27   Si (apfu) 2.746  2.727  2.758  2.651  2.720  Ti 0.015  0.027  0.019  0.088  0.066  Al 1.469  1.464  1.49  1.507  1.466  V 0  0  0  0  0  Fe 0.116  0.139  0.066  0.056  0.052  Mg 2.652  2.632  2.625  2.657  2.661  Ba 0.024  0.028  0.003  0.190  0.122  Na 0.028  0.076  0.120  0.067  0.079  K 0.938  0.917  0.916  0.653  0.708  Cl 0.020  0.020  0.004  0.021  0.013  F 0.280  0.155  0.094  0  0  OHcalc 1.700  1.826  1.903  1.979  1.987  Below detection limit (wt. %): CaO (0.02); Cr2O3, ZnO, MnO, CoO, NiO (0.03).    152  Table 4.7B: Average composition of phlogopite from Glencoe Island. Normalized on the basis of 12 anions per formula unit. Locality Glencoe main Glencoe main   Glencoe contact Glencoe gulch Lithology Spl-bearing pod Spl-bearing pod   Spl diopsidite Fo-Prg-Spl marble Sample 2A-SPL-1 2A-SPL-2 2C 2D n 4 σ 4  σ 3 σ 4  σ SiO2 (wt. %) 38.13  0.07  38.81  0.22  36.96  0.08  37.81  0.74  TiO2 0.92  0.02  0.28  0.02  0.30  0.01  1.02  0.12  Al2O3 17.30 0.05 16.78 0.11 18.90 0.12 16.60 0.24 V2O3 <0.02  0.04 0.01 <0.02  <0.02  FeO 1.51 0.03 1.27 0.04 2.30 0.01 0.45 0.02 MgO  24.12 0.12 24.99 0.05 23.33 0.09 25.39 0.39 BaO 2.21 0.10 1.40 0.08 0.92 0.04 4.04 0.73 Na2O 0.08 0.01 0.06 0.01 0.16 0.02 0.24 0.02 K2O 10.04 0.06 10.38 0.02 10.33 0.07 8.75 0.18 Cl 0.31 0.03 0.27 0.06 0.26 0.02 0.14              F <0.12  <0.12  <0.12  0.25 0.15 H2Ocalc 4.09  4.11   4.07  3.99   O=F,Cl  -0.07  -0.06  -0.06  -0.14   TOTAL  98.64  98.33  97.47  98.54  Si (apfu) 2.745  2.785  2.682  2.734  Ti 0.050  0.015  0.016  0.055  Al 1.468  1.419  1.616  1.415  V 0  0.002  0  0  Fe 0.091  0.076  0.140  0.027  Mg 2.589  2.673  2.524  2.737  Ba 0.062  0.039  0.026  0.114  Na 0.011  0.008  0.023  0.034  K 0.922  0.950  0.956  0.807  Cl 0.038  0.033  0.032  0.017  F 0  0  0  0.057  OHcalc 1.962  1.967  1.968  1.926  Below detection limit (wt. %): CaO (0.02); Cr2O3, ZnO, MnO, CoO, NiO (0.03).    153  Table 4.7C: Average composition of phlogopite from the Soper Lake mine, Kimmirut area. Normalized on the basis of 12 anions per formula unit. Locality Soper Lk mine Soper Lk mine Lithology Fo-Carb-Phl-Spl rock Fo-Spl marble Sample 3D-1 3D-M1 n 6 σ 4 σ SiO2 (wt.%) 39.80 0.35 39.11 0.34 TiO2 1.20 0.02 0.94 0.01 Al2O3 15.73 0.14 16.12 0.02 V2O3 < 0.02  < 0.02  FeO 0.59 0.05 0.70 0.03 MgO 26.09 0.14 25.82 0.24 BaO 0.16 0.03 0.28 0.03 Na2O 0.99 0.02 1.26 0.05 K2O 9.97 0.07 9.44 0.03 Cl 0.11 0.01 0.14 0.01 F 1.52 0.06 1.32 0.05 H2Ocalc 3.50  3.55  O=F,Cl -0.66  -0.59  TOTAL 99.00  98.09  Si (apfu) 2.807  2.783  Ti 0.064  0.050  Al 1.307  1.352  V 0  0  Fe 0.035  0.042  Mg 2.743  2.739  Ba 0.004  0.008  Na 0.135  0.174  K 0.897  0.857  Cl 0.013  0.017  F 0.339  0.297  OHcalc 1.648  1.686  Below detection limit (wt. %): CaO (0.02); Cr2O3, ZnO, MnO, CoO, NiO (0.03).   154  Table 4.7D: Average composition of phlogopite from Qila, Kimmirut area. Normalized on the basis of 12 anions per formula unit. Locality Qila Qila Qila Lithology Spl-bearing calc-silicate Prg-Cal rock Spl-bearing marble Sample 3E-1 3E-2 3E-3-A n 6 σ 4 σ 3 σ SiO2 (wt.%) 41.58 0.26 39.42 0.21 39.46 0.15 TiO2 0.60 0.07 0.29 0.01 0.39 0.01 Al2O3 13.85 0.27 16.19 0.06 16.33 0.08 V2O3 < 0.02  < 0.02  < 0.02  FeO 0.89 0.05 0.51 0.04 0.57 0.01 MgO 26.35 0.24 26.30 0.04 25.69 0.11 BaO 0.51 0.03 0.95 0.05 0.8 0.06 Na2O 0.11 0.02 0.72 0.02 0.63 0.03 K2O 9.90 0.13 9.63 0.08 9.8 0.01 Cl 0.07 0.01 0.14 0 0.13 0.01 F 1.58 0.13 0.97 0.06 0.42 0.16 H2Ocalc 3.46  3.72  3.97  O=F,Cl -0.69  -0.44  -0.21  TOTAL 98.21  98.40  97.98  Si (apfu) 2.947  2.802  2.813  Ti 0.032  0.016  0.021  Al 1.157  1.356  1.372  V 0  0  0  Fe 0.053  0.03  0.034  Mg 2.784  2.787  2.730  Ba 0.014  0.026  0.022  Na 0.015  0.099  0.087  K 0.895  0.873  0.891  Cl 0.008  0.017  0.016  F 0.354  0.218  0.095  OHcalc 1.637  1.765  1.890  Below detection limit (wt. %): CaO (0.02); Cr2O3, ZnO, MnO, CoO, NiO (0.03).   155  Table 4.7E: Average composition of phlogopite from Trailside, Kimmirut area. Normalized on the basis of 12 anions per formula unit. Locality Trailside Trailside Trailside Lithology Spl-bearing silicate rock Spl-bearing silicate rock Cal-Phl-Spl rock Sample 3F-1 3F-1 3F-2 n 3 σ 4 σ 4 σ SiO2 (wt.%) 36.96 0.36 38.88 0.11 39.20 0.15 TiO2 0.09 0.01 0.32 0.01 0.20 0.01 Al2O3 22.37 0.47 18.12 0.10 18.29 0.05 V2O3 < 0.02  < 0.02  < 0.02  FeO 1.33 0.04 1.25 0.03 1.18 0.02 MgO 23.24 0.02 24.40 0.07 24.81 0.08 BaO 0.10 0.05 0.29 0.02 0.35 0.03 Na2O 0.32 0.03 0.32 0.02 0.44 0 K2O 10.53 0.07 10.49 0.05 10.25 0.03 Cl < 0.01  0.02 0.01 0.07 0.01 F < 0.12  < 0.12  < 0.12  H2Ocalc 4.26  4.21  4.24  O=F,Cl 0  0  -0.02  TOTAL 99.20  98.30  99.01  Si (apfu) 2.599  2.763  2.762  Ti 0.005  0.017  0.011  Al 1.854  1.517  1.519  V 0  0  0  Fe 0.078  0.074  0.070  Mg 2.436  2.585  2.606  Ba 0.003  0.008  0.010  Na 0.044  0.044  0.060  K 0.945  0.951  0.921  Cl 0  0.002  0.008  F 0  0  0  OHcalc 2.000  1.998  1.992  Below detection limit (wt. %): CaO (0.02); Cr2O3, ZnO, MnO, CoO, NiO (0.03).    156  Table 4.8: Average composition of humite from Soper Falls. Normalized on the basis of 3 Si atoms per formula unit. Locality Soper Falls Lithology Humitite Sample 3B n 5     SiO2 (wt.%) 36.72     TiO2 0.43      FeO 1.18      MnO 0.05      MgO 57.21        F 4.70 H2O (calc) 2.72    TOTAL 103.07 Si (apfu) 3 Ti 0.026 Fe 0.081 Mn 0.003 Mg 6.968 F 1.214 OH (calc) 0.786 Below detection limit (wt. %): ZnO, Cr2O3, CoO, NiO (0.03); Al2O3, V2O3 (0.02); Cl (0.01).         157  Table 4.9: Average composition of scapolite from spinel-bearing calc-silicate rock at Qila. Normalized to Al+Si=12. Locality Qila Lithology Spl-bearing calc-silicate Sample 3E-1 n 3 SiO2 (wt.%) 47.76 Al2O3 26.98 MgO 0.04 CaO 16.52 Na2O 4.14 K2O 0.34 Cl 0.86 O=Cl -0.19 TOTAL 96.45 Si (apfu) 7.204 Al 4.796 Mg 0.009 Ca 2.670 Na 1.211 K 0.065 Me% 69 Below detection limit (wt. %): TiO2 (0.02); V2O3, Cr2O3, FeO, CoO, NiO, MnO (0.03); ZnO (0.04); BaO (0.06).     158  Table 4.10: Average composition of muscovite from spinel-bearing silicate rock at Trailside. Normalized to 12 anions per formula unit. Locality Trailside  Lithology Spl-bearing silicate rock Sample 3F-1  n 3 σ SiO2 (wt.%) 45.83           2.39  Al2O3  35.83           2.93  FeO  0.08           0.04  MgO  1.65           1.18  CaO  0.73           0.19  BaO  0.09           0.03  Na2O  0.25           0.10  K2O  8.75           0.32  Cl  0.16           0.17  H2O (calc)  4.43   O=F,Cl  -0.04  TOTAL          97.76   Si (apfu) 3.071  Al 2.829  Fe 0.004  Mg 0.165  Ca 0.052  Ba 0.002  Na 0.032  K 0.748  Cl 0.018  OH (calc) 1.982  Below detection limit (wt. %): TiO2 (0.01); V2O3 (0.02); Cr2O3, ZnO, MnO, CoO, NiO (0.03); F (0.12).  159  Table 4.11A: Average composition of calcite associated with spinel from Markham Bay and Glencoe Island. Locality Spinel Island Unnamed Island Glencoe main Glencoe main Glencoe main Glencoe beach Glencoe beach Glencoe contact Glencoe gulch Lithology Spl diopsidite Spl-bearing metasomatite Spl-bearing pod Spl-bearing pod Spl-bearing pod Spl-bearing marble Spl-bearing marble Spl diopsidite Fo-Prg-Spl marble Sample 1A-SPL 1B 2A-SPL-1 2A-SPL-1 2A-SPL-3 2B 2B 2C 2D n 4 3 2 2 3 2 2 3 3 FeO (wt.%) < 0.03 0.17 0.17 0.14 0.17 0.27 < 0.03 0.21 0.05 MnO < 0.03 0.03 0.06 0.05 < 0.03 0.09 < 0.03 0.06 < 0.03 MgO 0.03 1.53 2.36 2.44 2.96 2.95 0.35 1.84 2.62 CaO 57.43 54.40 53.19 53.63 52.45 52.62 56.01 53.54 53.29 CO2 43.07 43.56 43.75 43.65 43.87 43.78 43.34 43.70 43.74 TOTAL 100.56 99.69 99.53 99.91 99.47 99.71 99.81 99.35 99.72  Table 4.11B: Average composition of calcite associated with spinel from Kimmirut, Waddell Bay, and the Hall Peninsula. Locality Soper River Soper River Soper Falls Soper Lk mine Soper Lk mine Qila Qila Trailside Trailside Waddell Bay Hall Pen. Lithology Spl-bearing marble Spl-bearing marble Humitite Fo-Carb-Phl-Spl rock Fo-Spl marble Prg-Cal rock Spl-bearing marble Cal-Phl-Spl rock Cal-Phl-Spl rock Cal-Di-Phl-Spl rock Spl-bearing marble Sample 3A-2 3A-1 3B 3D-1 3D-M1 3E-2 3E-3-A 3F-2 3F-2 4 5 n 3 3 3 3 3 3 3 3 3 3 3 FeO (wt.%) 0.03 0.03 0.04 < 0.03 0.09 0.07 0.06 0.20 0.20 0.06 0.09 MnO 0.04 0.04 0.04 0.03 0.05 0.04 < 0.03 0.04 0.04 < 0.03 0.04 MgO 0.35 1.06 2.50 0.48 2.26 3.24 1.34 3.02 3.45 1.34 2.44 CaO 56.61 55.53 53.64 56.61 53.86 52.54 55.23 52.77 52.58 54.54 53.32 CO2 43.21 43.39 43.67 43.20 43.63 43.84 43.42 43.77 43.77 43.58 43.74 TOTAL 100.24 100.05 99.89 100.32 99.89 99.73 100.07 99.8 100.04 99.53 99.63 160  Table 4.12A: Average composition of dolomite and dolomite exsolution in calcite associated with spinel from Baffin Island. Locality Glencoe main Glencoe gulch Soper River Soper River Soper River Soper Falls Lithology Spl-bearing pod Fo-Prg-Spl marble Spl-bearing marble Spl-bearing marble Spl-bearing marble Humitite Sample 2A-SPL-1 2D 3A-2 3A-1 3A-1 3B Detail Exsolution Exsolution Exsolution  Di-Dol pseudomorph Exsolution n 2 3 3 3 2 3 FeO (wt.%) 0.74 0.18 0.25 0.23 0.26 0.24 MnO 0.07 0.03 < 0.03 0.05 0.04 0.05 MgO 20.40 21.13 21.56 21.08 20.89 21.28 CaO 31.11 31.46 30.86 30.78 30.68 31.09 CO2 46.89 46.89 46.97 47.06 47.11 46.94 TOTAL 99.21 99.69 99.66 99.20 98.98 99.60  Table 4.12B: Average composition of dolomite and dolomite exsolution in calcite associated with spinel from Baffin Island. Locality Soper Lk mine Soper Lk mine Qila (area) Qila Qila Hall Pen. Lithology Fo-Spl marble Fo-Carb-Phl-Spl rock Dolomitic marble Spl-bearing marble Spl-bearing calc-silicate Spl-bearing marble Sample 3D-M1 3D-1 3E-M1 3E-3-A 3E-1 5 Detail Exsolution Exsolution  Exsolution  Exsolution n 2 3 3 2 3 3 FeO (wt.%) 0.39 0.34 0.18 0.36 0.55 0.36 MnO 0.07 0.07 < 0.03 0.06 0.06 0.03 MgO 21.06 21.03 21.44 20.82 20.58 21.19 CaO 30.82 30.77 30.54 31.01 30.74 30.62 CO2 46.99 47.02 47.10 46.99 47.03 47.05 TOTAL 99.33 99.23 99.28 99.24 98.96 99.25 161  Table 4.13: Average composition of pyrrhotite and pyrite associated with spinel on Baffin Island. Species Pyrrhotite Pyrrhotite Pyrrhotite Pyrrhotite Pyrite Locality Glencoe main Soper River Soper Falls Soper Lk camp Trailside Lithology Spl-bearing pod Spl-bearing marble Humitite Spl-bearing marble Spl-bearing silicate rock Sample 2A-SPL-1 3A-1 3B 3C 3F-1 (after Spl) n 4 3 3 5 1 Fe (wt. %) 60.99 61.79 63.19 60.32 43.06 Co 0.07 0.11 0.06 0.09 1.79 Ni 0.21 0.10 < 0.03 0.04 < 0.03 S 38.59 38.03 36.81 38.62 52.75 TOTAL 99.86 100.05 100.08 99.09 97.61 Below detection limit (wt. %): Mn (0.03), Zn (0.04). 162   Table 4.14A: Major element composition of whole rock samples from Spinel Island, Glencoe Island, and the Hall Peninsula. Locality Spinel Island Spinel Island Spinel Island Spinel Island Glencoe main Glencoe main Glencoe main Glencoe main Glencoe main Hall Pen. Lithology Monzogranite Syenogranite Phl marble Spl diopsidite Spl-bearing pod Spl-bearing pod (w/ Po) Serpentine marble Sulphide-rich diopsidite Psammite Spl-bearing marble Sample 1A-MG 1A-SG 1A-M 1A-SPL 2A-SPL-1 2A-SPL-2 2A-M 2A-DI 2A-PS 5 Wt. %                     SiO2 72.5 75.1 7.96 45.5 38.1 34.9 20.8 46.8 71 3.93 Al2O3 14.75 13.6 1.67 10.4 17.85 12.6 1.48 4.91 11.8 1.13 TiO2 0.31 0.03 0.07 0.13 0.28 0.3 0.04 0.2 0.34 0.05 Cr2O3 <0.01 <0.01 <0.01 <0.01 0.01 0.01 <0.01 0.01 0.01 <0.01 MnO 0.01 0.01 0.05 0.05 0.03 0.04 0.06 0.04 0.04 0.04 Fe2O3 1.24 0.54 0.59 2.02 1.11 1.8 0.94 5.36 4.89 0.43 MgO 0.48 0.19 10.8 17.55 11 17.35 21.7 16.25 3.7 8.02 CaO 0.81 0.47 41.9 22.3 22.2 21 26.6 21.9 6.18 47.6 SrO 0.02 0.01 0.01 <0.01 0.07 <0.01 0.01 <0.01 0.01 0.01 Na2O 3.77 1.85 0.06 0.09 0.44 0.07 0.05 0.05 0.9 0.03 K2O 7.07 9.07 1.1 0.23 1.02 2.29 0.81 0.65 0.49 0.03 BaO 0.14 0.12 0.12 0.02 0.32 0.56 0.17 0.1 0.06 0.02 P2O5 0.08 0.05 <0.01 0.01 0.23 0.18 0.03 0.02 0.03 <0.01 LOI 0.54 0.33 36.2 1.22 6.86 7.57 28.1 4.12 2.26 39.9 Total 101.72 101.37 100.53 99.52 99.52 98.67 100.79 100.41 101.71 101.19 C 0.08 0.03 9.83 0.23 1.77 2 5.87 2.11 0.57 10.9 S <0.01 0.01 0.04 0.01 0.1 0.25 0.09 1.97 1.33 <0.01    163   Table 4.14B: Major element composition of whole rock samples from Soper River, Soper Lake, and Soper Falls, Kimmirut area. Locality Soper River Soper River Soper River Soper Falls Soper Lk camp Soper Lk mine Soper Lk mine Soper Lk mine Lithology Spl-bearing marble Lapis lazuli Lapis lazuli Humitite Spl-bearing marble Fo-Carbonate rock Fo-Spl marble Phl marble Sample 3A-1 3A-LAPIS-1 3A-LAPIS-2 3B 3C 3D-2 3D-M1 3D-M2 Wt. %                 SiO2 2.2 35.7 42.1 29.8 5.94 33.4 5.95 6.24 Al2O3 0.39 10.85 12.25 8.11 1.26 0.23 1.32 1.35 TiO2 0.02 0.16 0.18 0.3 0.12 0.01 0.01 0.07 Cr2O3 <0.01 0.01 0.01 0.01 <0.01 <0.01 <0.01 <0.01 MnO 0.05 0.02 0.02 0.05 0.03 0.08 0.06 0.04 Fe2O3 0.31 0.39 0.29 1.55 0.83 2.58 0.75 0.63 MgO 16.75 8.16 10.2 46.4 11.85 43.2 10.75 7.56 CaO 37.5 24.1 19.25 3.79 42.8 6.08 44.5 46.3 SrO <0.01 <0.01 <0.01 <0.01 0.01 0.01 0.02 0.02 Na2O 0.04 6.33 6.96 0.06 0.03 0.04 0.06 0.08 K2O 0.01 0.45 0.46 <0.01 <0.01 0.11 0.13 0.95 BaO <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 0.02 P2O5 0.01 0.01 0.03 0.02 <0.01 0.04 0.02 0.04 LOI 44.2 11.5 5.5 6.99 37.4 13.6 37.8 38.3 Total 101.48 97.68 97.25 97.08 100.27 99.38 101.37 101.6 C 12.15 2.86 1.36 1.45 10.2 1.92 10.35 10.3 S 0.01 0.98 1.27 0.02 0.11 0.02 0.03 0.08    164   Table 4.14C: Major element composition of whole rock samples from Qila and Trailside, Kimmirut area. Locality Qila Qila Qila Qila Qila (area) Trailside Trailside Trailside Trailside (area) Trailside (area) Trailside (area) Trailside (area) Lithology Calc-silicate* Prg-Cal rock Spl-bearing marble Phl-richer, Spl-poor marble Marble Spl-bearing silicate rock (Spl-poor) Cal-Phl-Spl rock Diopsidite Marble Marble Diopsidite Diopsidite Sample 3E-1 3E-2 3E-3-A 3E-3-B 3E-M2 3F-1 3F-2 3F-4 3F-M1 3F-M2 3F-CS2 3F-CS3 Wt. %                         SiO2 40.9 27.4 7.37 7.85 5.28 32.6 21.6 48.6 8.86 3.84 49.6 46.6 Al2O3 14.4 10.75 0.95 1.04 0.73 21 13.65 1.9 1.88 0.92 4.39 7.31 TiO2 0.15 0.12 <0.01 0.02 0.02 0.08 0.14 0.03 0.06 0.02 0.1 0.26 Cr2O3 0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 MnO 0.03 0.02 0.04 0.03 0.04 0.02 0.03 0.04 0.03 0.03 0.03 0.03 Fe2O3 0.82 0.54 0.56 0.33 0.53 0.44 1.14 0.88 0.72 0.31 1.51 1.39 MgO 15 13.75 11.5 8.16 7.33 8.08 16.35 16.3 9.02 3.06 16.2 18.15 CaO 12.9 26 43 46.3 48 15.6 21.4 23.5 43.4 50.1 19.95 16.55 SrO <0.01 0.01 0.02 0.02 0.02 0.03 0.02 <0.01 0.03 0.04 <0.01 <0.01 Na2O 1.58 1.41 0.04 0.06 0.03 0.69 0.24 0.63 0.07 0.16 0.9 1.06 K2O 1.86 1.38 0.07 0.42 0.5 5.07 5.65 0.36 1.13 0.32 0.88 3.13 BaO 0.1 0.08 0.01 0.04 0.02 0.17 0.2 0.01 0.02 0.02 0.01 0.03 P2O5 0.65 0.08 0.03 0.03 0.03 0.01 0.02 0.01 0.01 0.07 <0.01 0.02 LOI 10.8 16.8 36.4 37.1 39.4 18.1 20.1 8.79 36.3 41.2 5.52 5.37 Total 99.2 98.34 99.99 101.4 101.93 101.89 100.54 101.05 101.53 100.09 99.09 99.9 C 2.35 4.66 9.9 10.05 10.75 3.36 4.94 2.32 9.66 11.2 1.24 1.05 S 0.01 0.01 <0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 165  Table 4.15A: Trace element concentrations (µg/g) of whole rock samples from Spinel Island, Glencoe Island, and the Hall Peninsula. Below detection limit: Ge (< 5 µg/g). Locality Spinel Island Spinel Island Spinel Island Spinel Island Glencoe main Glencoe main Glencoe main Glencoe main Glencoe main Hall Pen. Lithology Monzogranite Syenogranite Phl marble Spl diopsidite Spl-bearing pod Spl-bearing pod (w/ Po) Serpentine marble Sulphide-rich diopsidite Psammite Spl-bearing marble Sample 1A-MG 1A-SG 1A-M 1A-SPL 2A-SPL-1 2A-SPL-2 2A-M 2A-DI 2A-PS 5 Li 10 <10 <10 10 10 30 <10 10 20 <10 B n.d. n.d. n.d. n.d. 21 n.d. n.d. n.d. n.d. n.d. F n.d. n.d. n.d. n.d. 1070 n.d. n.d. n.d. n.d. n.d. Cl n.d. n.d. n.d. n.d. 820 n.d. n.d. n.d. n.d. n.d. Sc 1 <1 2 3 9 12 1 7 4 1 V 19 <5 10 14 53 43 17 457 29 7 Cr 10 10 10 10 60 100 10 50 50 10 Co 2 <1 2 2 2 4 3 16 7 <1 Ni 1 <1 2 <1 11 23 4 149 35 <1 Cu 1 <1 2 1 6 17 4 55 62 <1 Zn 14 7 <2 98 230 308 242 231 67 2 Ga 18.4 18.3 2.5 20.2 17.5 19 2.2 8.7 17.6 1.4 As 0.4 0.4 0.2 0.3 0.8 0.7 <0.1 2.7 0.8 <0.1 Se <0.2 <0.2 0.2 <0.2 0.3 1.2 0.2 2.8 0.3 <0.2 Rb 263 372 32.4 20.9 33.4 76.4 29.7 33.7 15 1.3 Sr 167.5 144.5 86.6 18.1 614 49.2 83.3 26.7 87.1 93.3 Y 2.5 2.8 6.6 14.1 11.2 9.4 7.2 16.4 9.3 2.7 Zr 120 3 43 116 284 212 11 86 209 17 Nb 5.8 0.8 2.5 2.2 4.6 9.1 2 8.6 11.1 2.5 Mo <1 1 1 <1 1 4 2 48 3 <1 Ag <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 0.7 <0.5 <0.5 Cd <0.5 <0.5 <0.5 <0.5 0.9 1.2 0.9 1.5 0.6 <0.5 In <0.005 <0.005 0.009 0.007 <0.005 0.007 <0.005 0.014 0.011 0.005 Sn <1 <1 1 4 2 2 <1 2 1 <1 Sb <0.05 <0.05 0.12 <0.05 0.07 0.1 <0.05 0.11 <0.05 <0.05 Te <0.01 <0.01 0.01 <0.01 0.01 <0.01 <0.01 0.02 0.02 <0.01 Cs 0.46 1.05 0.33 0.56 1.18 3.08 1.22 1.09 0.17 0.11 Ba 1265 1075 1100 173.5 2970 5300 1540 937 565 189.5 La 20.2 17.6 10.5 15.5 10.2 6.9 14 8 29.4 4.1 Ce 30.9 26.4 19.9 47.1 19.8 16 21.8 21.2 47.6 8.4 Pr 3.27 2.76 2.36 7.31 2.87 2.31 2.61 3.27 5.23 1.02 Nd 10.3 8.9 8.3 27.5 10.6 9.5 8.7 12.9 15.9 3.9 Sm 1.66 1.39 1.5 5.44 2.45 2.26 1.37 2.69 2.81 0.57 Eu 0.88 1.01 0.27 0.43 0.31 0.23 0.23 0.4 0.62 0.09 Gd 0.82 1.03 1.24 3.6 1.81 1.71 1.05 2.37 1.74 0.6 Tb 0.12 0.11 0.17 0.52 0.32 0.26 0.13 0.39 0.26 0.09 Dy 0.54 0.57 0.97 2.74 1.82 1.51 0.84 2.44 1.31 0.5 Ho 0.08 0.09 0.2 0.53 0.42 0.32 0.18 0.56 0.31 0.09 Er 0.2 0.3 0.67 1.24 1.17 1 0.64 1.39 0.9 0.28 Tm 0.05 0.04 0.1 0.16 0.19 0.17 0.08 0.21 0.14 0.04 Yb 0.22 0.15 0.58 1.04 1.07 1.14 0.53 1.33 1.06 0.3 Lu 0.03 0.02 0.1 0.16 0.19 0.17 0.09 0.21 0.15 0.05 Hf 2.9 <0.2 1.1 3.4 7.6 6.9 0.3 2.1 5.6 0.5 Ta 0.3 0.1 0.2 0.4 0.7 0.9 0.1 0.4 0.9 0.2 W <1 1 <1 <1 <1 1 1 5 1 <1 Re <0.001 <0.001 <0.001 <0.001 0.003 0.006 0.002 0.116 0.002 <0.001 Hg <0.005 0.005 <0.005 0.006 <0.005 0.007 0.007 <0.005 0.01 <0.005 Tl 0.06 0.03 0.02 0.08 0.34 0.91 0.15 0.48 0.05 <0.02 Pb 21 32 2 24 32 42 12 13 7 2 Bi 0.01 0.01 0.04 0.01 0.05 0.16 0.01 0.3 0.2 0.03 Th 4 4.19 2.64 101.5 25.6 16.85 0.84 2 15.05 1.99 U 0.86 2.24 0.18 47.8 32.7 27.1 0.67 1.51 2.92 0.34   166  Table 4.15B: Trace element concentrations (µg/g) of whole rock samples from Soper River, Soper Lake, and Soper Falls, Kimmirut area. Below detection limit (µg/g): Ge (5), Cd (0.5). Locality Soper River Soper River Soper River Soper Falls Soper Lk camp Soper Lk mine Soper Lk mine Soper Lk mine Lithology Spl-bearing marble Lapis lazuli Lapis lazuli Humitite Spl-bearing marble Fo-Carbonate rock Fo-Spl marble Phl marble Sample 3A-1 3A-LAPIS-1 3A-LAPIS-2 3B 3C 3D-2 3D-M1 3D-M2 Li <10 <10 <10 <10 <10 10 <10 <10 B 13 n.d. 17 304 n.d. n.d. 129 n.d. F n.d. n.d. 560 >20000 n.d. n.d. 510 n.d. Cl n.d. n.d. 3170 250 n.d. n.d. 390 n.d. Sc <1 4 4 <1 1 1 1 1 V 5 30 38 35 10 5 8 11 Cr 10 30 50 40 10 10 10 10 Co 1 <1 <1 3 1 6 2 2 Ni <1 <1 <1 <1 1 7 <1 3 Cu 6 <1 <1 1 5 1 3 4 Zn <2 4 10 5 3 14 4 5 Ga 0.5 13.3 10.2 13.4 1.5 0.7 2.3 1.9 As 0.1 5.7 19.4 0.9 <0.1 0.3 0.2 <0.1 Se <0.2 1 0.9 <0.2 0.4 <0.2 <0.2 0.3 Rb <0.2 2.4 5.9 0.2 0.5 3.5 4.6 59.8 Sr 26.7 49.5 31.8 18.9 86.6 105.5 192 203 Y 0.9 6 3.9 <0.5 4.2 1.4 4.6 4.3 Zr 26 91 105 16 40 33 16 35 Nb 0.2 1.5 4.3 1.9 0.8 0.9 0.9 3.1 Mo <1 5 22 <1 <1 <1 <1 <1 Ag <0.5 <0.5 <0.5 <0.5 <0.5 <0.5 0.5 <0.5 In <0.005 <0.005 <0.005 0.005 0.007 0.008 0.005 0.007 Sn <1 1 1 <1 <1 <1 <1 <1 Sb <0.05 <0.05 0.09 <0.05 <0.05 <0.05 <0.05 <0.05 Te <0.01 <0.01 0.01 <0.01 0.01 <0.01 0.01 <0.01 Cs 0.01 0.15 0.23 0.02 0.02 0.28 0.65 6.4 Ba 6.4 15.6 15 6.8 19.2 18.9 42.4 205 La 1.1 3.7 2.2 <0.5 7.1 2.7 8.6 8.8 Ce 2.1 9.6 5.6 0.9 15.4 5.1 17.4 17.2 Pr 0.21 1.28 0.73 0.13 1.47 0.57 1.85 1.67 Nd 0.9 5.3 3.4 0.5 5.7 2.1 6.1 5.2 Sm 0.12 1.23 0.73 <0.03 1.04 0.34 1.23 0.87 Eu 0.03 0.18 0.13 <0.03 0.17 0.06 0.25 0.18 Gd 0.16 1.1 0.77 0.07 0.92 0.35 0.88 0.69 Tb 0.02 0.15 0.11 0.01 0.14 0.03 0.12 0.11 Dy 0.12 1.04 0.75 0.05 0.74 0.21 0.77 0.61 Ho 0.02 0.2 0.16 0.01 0.15 0.05 0.14 0.13 Er 0.1 0.69 0.45 <0.03 0.39 0.14 0.42 0.38 Tm 0.02 0.09 0.07 0.01 0.07 0.03 0.06 0.06 Yb 0.07 0.68 0.4 0.03 0.39 0.15 0.38 0.33 Lu 0.01 0.11 0.07 <0.01 0.06 0.02 0.06 0.05 Hf 0.7 2.6 3.1 0.7 0.9 1 0.4 0.9 Ta <0.1 0.3 0.4 0.2 0.3 <0.1 <0.1 0.2 W <1 1 5 1 <1 1 <1 <1 Re 0.001 0.001 0.003 <0.001 <0.001 <0.001 <0.001 <0.001 Hg <0.005 <0.005 0.006 <0.005 0.007 0.014 0.009 0.008 Tl <0.02 0.05 0.09 <0.02 <0.02 0.02 0.02 0.16 Pb <2 <2 3 <2 4 5 5 5 Bi <0.01 0.02 0.03 0.01 0.03 0.03 0.03 0.03 Th 0.1 1.21 0.92 0.11 1.47 0.52 3.5 3.97 U <0.05 0.93 1.09 0.11 0.79 0.52 3.72 2.45     167  Table 4.15C: Trace element concentrations (µg/g) of whole rock samples from Qila, Kimmirut area. Below detection limit (µg/g): Ge (5); Ag and Cd (0.5); Re (0.001). Locality Qila Qila Qila Qila Qila (area) Lithology Calc-silicate* Prg-Cal rock Spl-bearing marble Phl-richer, Spl-poor marble Marble Sample 3E-1 3E-2 3E-3-A 3E-3-B 3E-M2 Li 30 10 <10 <10 <10 B n.d. 95 n.d. n.d. n.d. F n.d. 6960 n.d. n.d. n.d. Cl n.d. 1160 n.d. n.d. n.d. Sc 5 4 1 1 1 V 13 13 <5 5 6 Cr 40 10 10 10 10 Co 29 9 15 10 3 Ni 47 11 25 12 4 Cu 1 2 1 1 2 Zn 2 2 2 <2 3 Ga 14.1 11 1.9 1.7 1 As 1.1 0.3 <0.1 0.2 <0.1 Se <0.2 0.2 <0.2 0.2 <0.2 Rb 15 12.4 1.6 12.4 20.5 Sr 50.9 98.6 177.5 170 191.5 Y 6.3 4 3.6 4.2 3.8 Zr 126 125 5 13 15 Nb 17.7 13.4 0.4 0.9 1.4 Mo <1 <1 3 3 <1 In 0.016 0.011 0.006 <0.005 0.006 Sn 4 4 <1 <1 <1 Sb <0.05 <0.05 <0.05 <0.05 <0.05 Te 0.01 <0.01 <0.01 <0.01 <0.01 Cs 0.3 0.24 0.04 0.38 0.62 Ba 892 664 87.4 345 165.5 La 1 1.6 3.9 4.2 5.8 Ce 2.7 3.1 7.5 7.9 10.7 Pr 0.42 0.39 0.84 0.94 1.16 Nd 1.9 1.7 3.5 3.3 4.5 Sm 0.63 0.41 0.64 0.58 0.77 Eu 0.05 0.06 0.09 0.08 0.11 Gd 0.65 0.64 0.59 0.61 0.64 Tb 0.16 0.1 0.08 0.11 0.1 Dy 0.92 0.64 0.46 0.64 0.67 Ho 0.24 0.15 0.12 0.14 0.12 Er 0.73 0.48 0.33 0.4 0.37 Tm 0.15 0.09 0.05 0.06 0.05 Yb 0.85 0.48 0.29 0.39 0.31 Lu 0.12 0.06 0.04 0.06 0.05 Hf 5 3.9 0.2 0.4 0.5 Ta 1.6 1.2 <0.1 0.1 0.1 W 1 1 <1 <1 <1 Hg 0.082 <0.005 <0.005 <0.005 <0.005 Tl 0.04 0.03 <0.02 <0.02 0.04 Pb 99 6 2 3 2 Bi 0.04 0.02 0.01 0.01 0.01 Th 84.2 3.26 3.78 4.58 2.31 U 257 8.21 3.38 3.49 1.51     168  Table 4.15D: Trace element concentrations (µg/g) of whole rock samples from Trailside, Kimmirut area. Below detection limit (µg/g): Ge (5); Ag and Cd (0.5); Re (0.001). Locality Trailside Trailside Trailside Trailside (area) Trailside (area) Trailside (area) Trailside (area) Lithology Spl-bearing silicate rock (Spl-poor) Cal-Phl-Spl rock Diopsidite Marble Marble Diopsidite Diopsidite Sample 3F-1 3F-2 3F-4 3F-M1 3F-M2 3F-CS2 3F-CS3 Li 50 30 <10 <10 <10 10 10 B 196 n.d. n.d. 30 n.d. n.d. n.d. F 2320 n.d. n.d. 1970 n.d. n.d. n.d. Cl 680 n.d. n.d. 410 n.d. n.d. n.d. Sc 1 1 4 2 1 6 3 V 7 17 16 10 6 33 22 Cr 10 20 10 10 10 30 20 Co 9 27 16 2 <1 5 8 Ni 10 35 11 2 <1 2 12 Cu 1 <1 1 1 1 1 1 Zn <2 15 3 3 <2 14 9 Ga 7.4 14.6 3.6 2.7 1.4 7.7 9.4 As 1.6 0.1 0.5 0.2 <0.1 0.4 0.1 Se 0.2 <0.2 <0.2 <0.2 0.2 0.2 <0.2 Rb 85.1 143 8.8 29.4 9 33.8 91.2 Sr 316 217 42.4 276 373 33.5 32.8 Y 5.4 7.3 1.9 6.3 4 1.2 0.9 Zr 19 24 77 19 14 119 89 Nb 6.6 14.5 0.7 3.1 0.9 3.3 10.9 Mo <1 <1 <1 <1 <1 <1 <1 In <0.005 <0.005 <0.005 0.006 0.007 0.006 0.007 Sn 1 1 1 1 <1 2 2 Sb 0.05 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 Te <0.01 <0.01 <0.01 0.01 <0.01 <0.01 0.01 Cs 1.96 3.88 0.28 0.77 0.19 0.99 3.34 Ba 1500 1800 72.3 175.5 152 74.4 279 La 2.6 3.8 0.5 5.8 6.1 <0.5 <0.5 Ce 5.6 8.3 1.2 12 12 1 0.8 Pr 0.68 0.97 0.19 1.48 1.26 0.16 0.11 Nd 3 4 0.9 5.6 4.9 0.7 0.5 Sm 0.63 0.94 0.3 1.29 1.1 0.18 0.08 Eu 0.09 0.1 0.03 0.12 0.15 0.03 0.03 Gd 0.71 0.96 0.32 1 0.89 0.19 0.14 Tb 0.11 0.17 0.05 0.15 0.11 0.04 0.02 Dy 0.74 1.09 0.38 1.02 0.64 0.22 0.17 Ho 0.2 0.24 0.07 0.22 0.13 0.05 0.04 Er 0.66 0.77 0.28 0.68 0.37 0.18 0.09 Tm 0.11 0.14 0.04 0.09 0.06 0.03 0.04 Yb 0.63 0.87 0.24 0.63 0.4 0.17 0.18 Lu 0.12 0.13 0.03 0.1 0.06 0.02 0.02 Hf 0.7 0.6 2.2 0.6 0.3 3.2 2.1 Ta 0.4 0.7 <0.1 0.2 0.1 0.2 0.9 W 1 1 <1 2 <1 <1 2 Hg <0.005 <0.005 <0.005 <0.005 <0.005 <0.005 0.008 Tl 0.09 0.2 0.03 0.06 <0.02 0.12 0.32 Pb 2 <2 <2 <2 <2 <2 <2 Bi 0.01 0.01 0.01 0.01 0.02 0.01 0.01 Th 3.75 2.39 0.7 3.41 3.92 0.47 0.37 U 1.32 0.95 0.73 1.18 2.15 0.59 0.56  169   Table 4.16A: Estimated protolith composition of metacarbonate samples from Markham Bay, Glencoe Island. Locality  Spinel Island Glencoe main Glencoe main Glencoe main Glencoe main Glencoe main Lithology  Phl marble Spl-bearing pod Spl-bearing pod (w/ Po) Serpentine marble Sulphide-rich diopsidite Psammite Sample  1A-M 2A-SPL-1 2A-SPL-2 2A-M 2A-DI 2A-PS Al/Si (weight)  0.24 0.53 0.41 0.08 0.12 0.19 Siliciclastic component  Sand-rich mud Muddy clay Mud Slightly muddy sand Muddy sand Mud-rich sand Siliciclastic Ca-Mga (wt. %) CaO 0.55 0.93 2.43 1.45 3.25 4.93  MgO 0.22 0.81 1.44 0.33 0.87 1.71 Original H2O estimate (wt. %)b  H2O 0.36 3.34 3.15 0.07 0.63 2.26 Original carbonatec CaO 41.35 21.27 18.57 25.15 18.65 1.25  MgO 10.58 10.19 15.91 21.37 15.38 1.99 Original carbonate CO2 44.00 27.82 31.95 43.08 31.42 3.15 Totald  108.73 123.92 126.45 115.93 130.31 106.19 Carbonate species (mol. %) Magnesite 0 0 16 15 13 55  Dolomite 36 67 84 85 87 45  Calcite 64 33 0 0 0 0 Original rock composition estimate (wt. %)e       Siliciclastic  Sand 4.9  2.4 21.3 39.0 52.9  Mud 6.7 16.1 44.6 1.2 10.0 41.0  Clay  35.8      Sili. Total 11.6 51.9 47 22.5 48.9 93.9 Carbonate  88.4 48.1 53 77.5 51.1 6.1 a Estimated based on reference averages and samples of modern sedimentary rocks (see text). b Assuming 5 wt. % water in shales and claystones. c Siliciclastic contribution subtracted from whole rock total. d Whole rock composition excluding volatiles and with the addition of the estimated original CO2, H2O, and where applicable, Cl. e Where the siliciclastic proportion is calculated assuming it contains all Al, Si, Ti, Cr, K, Na (except in lapis lazuli), and their calculated estimated contribution of Ca and Mg; and the carbonate proportion contains the remainder of Ca and Mg plus estimated original CO2.            170   Table 4.16B: Estimated protolith composition of metacarbonate samples from Hall Peninsula and part of the Kimmirut area. Locality  Beluga sapphire Soper River Soper Falls Soper Lk camp Soper Lk mine Soper Lk mine Hall Pen. Lithology  Calc-silicate Spl-bearing marble Humitite Spl-bearing marble Fo-Spl marble Phl marble Spl-bearing marble Sample  Belley et al. 2017 3A-1 3B 3C 3D-M1 3D-M2 5 Al/Si (weight)  0.35 0.2 0.31 0.24 0.25 0.24 0.33 Siliciclastic component  Sandy mud Mud-rich sand Sandy mud Sand-rich mud Sand-rich mud Sand-rich mud Sandy mud Siliciclastic Ca-Mga (wt. %) CaO 3.18 0.15 2.07 0.41 0.41 0.43 0.27  MgO 1.67 0.06 1.00 0.17 0.17 0.18 0.14 Original H2O estimate (wt. %)b  H2O 3.39 0.08 1.89 0.27 0.29 0.29 0.27 Original carbonatec CaO 13.88 37.35 1.72 42.39 44.09 45.87 47.33  MgO 8.48 16.69 45.4 11.68 10.58 7.38 7.88 Original carbonate CO2 20.15 47.54 50.93 46.02 46.15 44.06 45.75 Totald  117.96 104.91 142.93 109.27 110.04 107.73 107.31 Carbonate species (mol. %) Magnesite 0 0 97 0 0 0 0  Dolomite 85 62 3 38 33 22 23  Calcite 15 38 0 62 67 78 77 Original rock composition estimate (wt. %)e      Siliciclastic  Sand 10.8 1.6 7.9 3.5 3.2 3.9 1.3  Mud 53.1 1.5 23.5 4.8 5.1 5.7 4.6  Clay         Sili. Total 63.9 3.1 31.4 8.3 8.3 9.5 5.9 Carbonate  36.1 96.9 68.6 91.7 91.7 90.5 94.1 a Estimated based on reference averages and samples of modern sedimentary rocks (see text). b Assuming 5 wt. % water in shales and claystones. c Siliciclastic contribution subtracted from whole rock total. d Whole rock composition excluding volatiles and with the addition of the estimated original CO2, H2O, and where applicable, Cl. e Where the siliciclastic proportion is calculated assuming it contains all Al, Si, Ti, Cr, K, Na (except in lapis lazuli), and their calculated estimated contribution of Ca and Mg; and the carbonate proportion contains the remainder of Ca and Mg plus estimated original CO2. 171  Table 4.16C: Estimated protolith composition of metacarbonate samples from Soper River and Qila, Kimmirut area. Locality  Soper River Soper River Qila Qila Qila Qila Qila (area) Lithology  Lapis lazuli Lapis lazuli Calc-silicate* Prg-Cal rock Spl-bearing marble Phl-richer, Spl-poor marble Marble Sample  3A-LAPIS-1 3A-LAPIS-2 3E-1 3E-2 3E-3-A 3E-3-B 3E-M2 Al/Si (weight)  0.34 0.33 0.4 0.44 0.15 0.15 0.16 Siliciclastic component  Sandy mud Sandy mud Mud Clay Muddy sand Muddy sand Muddy sand Siliciclastic Ca-Mg-Naa (wt. %) CaO 2.48 2.93 2.84 1.84 0.51 0.55 0.37  MgO 1.29 1.48 1.65 1.16 0.15 0.17 0.11  Na2O 0.52 0.59      Original H2O estimate (wt. %)b  H2O 2.61 2.91 3.58 2.67 0.15 0.17 0.12 Original carbonatec CaO 21.62 16.32 10.06 24.16 42.49 45.75 47.63  MgO 6.87 8.72 13.35 12.59 11.35 7.99 7.22 Original carbonate CO2 24.46 22.34 22.46 32.71 45.73 44.64 45.26 Original evaporited Na 4.31 4.73       Cl 6.65 7.29      Totale  119.38 123.92 114.45 116.93 109.47 109.12 107.92 Carbonate species (mol. %) Magnesite 0 0 46 0 0 0 0  Dolomite 44 74 54 72 37 24 21  Calcite 56 26 0 28 63 76 79 Original rock composition (wt. %)f        Siliciclastic Sand 8 10.6 4.2  6.3 6.7 4.7  Mud 38 41 55.5 39.1 2.7 3.1 2.4  Clay    1.4     Kaol-rich clay         Sili. Total 46 51.5 59.6 40.5 9 9.7 7.1 Carbonate  44.7 38.7 40.4 59.5 91 90.3 92.9 Halite  9.3 9.8      a Estimated based on reference averages and samples of modern sedimentary rocks (see text). b Assuming 5 wt. % water in shales and claystones. c Siliciclastic contribution subtracted from whole rock total. d Siliciclastic contribution subtracted from whole rock total, for lapis lazuli only, assuming all excess Na is halite. e Whole rock composition excluding volatiles and with the addition of the estimated original CO2, H2O, and where applicable, Cl. f Where the siliciclastic proportion is calculated assuming it contains all Al, Si, Ti, Cr, K, Na (except in lapis lazuli), and their calculated estimated contribution of Ca and Mg; and the carbonate proportion contains the remainder of Ca and Mg plus CO2.    172  Table 4.16D: Estimated protolith composition of metacarbonate samples from Trailside, Kimmirut area. Locality  Trailside Trails. (area) Trails. (area) Trailside Trailside Trailside (area) Trailside (area) Lithology  Spl-bearing silicate rock (Spl-poor) Marble Marble Cal-Phl-Spl rock Diopsidite Diopsidite Diopsidite Sample  3F-1 3F-M1 3F-M2 3F-2 3F-4 3F-CS2 3F-CS3 Al/Si (weight)  0.73 0.24 0.27 0.72 0.04 0.1 0.18 Siliciclastic component  Kaolinite-rich clay Sand-rich mud Sand-rich mud Kaolinite-rich clay Sand Slightly muddy sand Mud-rich sand Siliciclastic Ca-Mga (wt. %) CaO 0.32 0.62 0.27 0.21 2.79 3.45 3.24  MgO 0.45 0.25 0.12 0.3 0.59 0.86 1.08 Original H2O estimate (wt. %)b  H2O 2.93 0.4 0.21 1.93 0 0.42 1.35 Original carbonatec CaO 15.28 42.78 49.83 21.19 20.71 16.5 13.31  MgO 7.63 8.77 2.94 16.05 15.71 15.34 17.07 Original carbonate CO2 20.32 43.15 42.32 34.16 33.41 29.71 29.08 Totale  107.05 108.79 101.43 116.54 125.68 123.71 124.97 Carbonate species (mol. %) Magnesite 0 0 0 5 5 23 44  Dolomite 69 29 8 95 95 77 56  Calcite 31 71 92 0 0 0 0 Original rock composition (wt. %)f        Siliciclastic Sand  5.4 2.1  44.4 43.5 31.2  Mud  7.5 4   6.7 21.2  Clay 7.2   7.3     Kaol-rich clay 52.3   31.3     Sili. Total 59.5 12.9 6.1 38.6 44.4 50.2 52.4 Carbonate  40.5 87.1 93.9 61.4 55.6 49.8 47.6 Halite         a Estimated based on reference averages and samples of modern sedimentary rocks (see text). b Assuming 5 wt. % water in shales and claystones. c Siliciclastic contribution subtracted from whole rock total. d Siliciclastic contribution subtracted from whole rock total, for lapis lazuli only, assuming all excess Na is halite. e Whole rock composition excluding volatiles and with the addition of the estimated original CO2, H2O, and where applicable, Cl. f Where the siliciclastic proportion is calculated assuming it contains all Al, Si, Ti, Cr, K, Na (except in lapis lazuli), and their calculated estimated contribution of Ca and Mg; and the carbonate proportion contains the remainder of Ca and Mg plus CO2.  173  Table 4.17A: “Expected” trace element composition of metasediment samples calculated based on mixing of sedimentary rock averages and comparison to actual values metasediments (calculated using averages for shale/clay, sandstone, and limestone and corrected for mass loss due to devolatilization; see text). Samples from Markham Bay, Glencoe Island, Hall Peninsula and part of the Kimmirut area. Locality Spinel Island Glencoe main Glencoe main Glencoe main Glencoe main Glencoe main Soper River Soper Falls Soper Lk camp Soper Lk mine Soper Lk mine Hall Pen. Lithology Phl marble Spl-bearing pod Spl-bearing pod (w/ Po) Serpentine marble Sulphide-rich diopsidite Psammite Spl-bearing marble Humitite Spl-bearing marble Fo-Spl marble Phl marble Spl-bearing marble Sample 1A-M 2A-SPL-1 2A-SPL-2 2A-M 2A-DI 2A-PS 3A-1 3B 3C 3D-M1 3D-M2 5 "Expected" concentration (µg/g)          Li 10 42 38 9 19 35 6 27 9 9 9 8 B 28 76 71 28 44 65 23 57 27 27 27 26 F 368 518 511 370 422 391 348 522 367 371 363 362 Cl 156 192 191 140 126 85 155 202 159 161 156 160 Sc 2 7 6 1 3 5 1 4 2 2 2 2 V 30 96 87 25 40 69 23 66 28 28 28 27 Cr 20 71 65 20 38 64 13 48 18 18 18 17 Mn 1100 1100 1100 1000 800 400 1100 1300 1100 1100 1100 1100 Fe 6600 23700 21600 6300 11800 20200 4500 16000 5900 6000 6100 5600 Co 2 13 11 0 3 9 0 7 1 1 1 1 Ni 26 73 67 20 27 44 22 52 25 26 25 25 Cu 8 39 35 6 13 28 5 24 7 7 8 7 Zn 26 63 59 23 32 45 22 48 25 25 25 24 Ga 7 22 20 7 13 20 5 15 6 6 6 6 Actual / Estimate            Li -- 0.2 0.8  0.5 0.6 -- -- -- -- -- -- B  0.3     0.6 5.3  4.8   F  2.1      > 38  1.4   Cl  4.3      1.2  2.4   Sc 1.2 1.3 1.9 0.8 2.8 0.8 -- -- 0.6 0.6 0.6 0.7 V 0.3 0.6 0.5 0.7 11.3 0.4 0.2 0.5 0.4 0.3 0.4 0.3 Cr 0.5 0.8 1.5 0.5 1.3 0.8 0.8 0.8 0.6 0.6 0.5 0.6 Mn 0.3 0.2 0.3 0.5 0.4 0.8 0.3 0.3 0.2 0.4 0.3 0.3 Fe 0.6 0.3 0.6 1 3.2 1.7 0.5 0.7 1 0.9 0.7 0.5 Co 1.3 0.2 0.4  5.7 0.8  0.4 0.8 1.7 1.5 -- Ni 0.1 0.2 0.3 0.2 5.6 0.8 -- -- 0 -- 0.1 -- Cu 0.2 0.2 0.5 0.7 4.4 2.2 1.2 0 0.7 0.4 0.5 -- Zn -- 3.6 5.2 10.5 7.3 1.5 -- 0.1 0.1 0.2 0.2 0.1 Ga 0.4 0.8 1 0.3 0.7 0.9 0.1 0.9 0.3 0.4 0.3 0.2 174   Table 4.17B: “Expected” trace element composition of metasediment samples calculated based on mixing of sedimentary rock averages and comparison to actual values metasediments (calculated using averages for shale/clay, sandstone, and limestone and corrected for mass loss due to devolatilization; see text). Samples from Qila and Trailside, Kimmirut area. Locality Qila Qila Qila Qila Qila (area) Trailside Trailside Trailside Trailside (area) Trailside (area) Trailside (area) Trailside (area) Lithology Calc-silicate* Prg-Cal rock Spl-bearing marble Phl-richer, Spl-poor marble Marble Spl-bearing silicate rock (Spl-poor) Cal-Phl-Spl rock Diopsidite Diopsidite Diopsidite Marble Marble Sample 3E-1 3E-2 3E-3-A 3E-3-B 3E-M2 3F-1 3F-2 3F-4 3F-CS2 3F-CS3 3F-M1 3F-M2 "Expected" concentration (µg/g)          Li 41 32 8 8 7 40 31 12 16 25 11 8 B 74 61 25 26 24 72 59 34 39 52 29 24 F 483 466 362 361 358 462 461 381 390 434 369 340 Cl 171 180 155 154 155 167 179 110 111 135 156 150 Sc 7 5 1 1 1 7 5 1 2 4 2 1 V 93 76 25 26 24 92 73 25 34 54 31 25 Cr 70 55 16 17 15 69 53 27 34 47 21 15 Mn 900 1100 1100 1100 1100 900 1100 800 800 900 1100 1100 Fe 23400 18400 5400 5600 5200 22900 17700 8100 10400 14900 6900 5200 Co 13 10 1 1 1 13 9 0.2 2 6 2 1 Ni 70 59 23 23 23 69 57 15 21 38 27 23 Cu 38 30 6 6 6 38 29 6 10 19 9 6 Zn 61 52 23 24 23 60 50 23 28 39 26 23 Ga 22 17 6 6 5 21 16 10 11 15 7 5 Actual / Estimate           Li 0.7 0.3 -- -- -- 1.2 1 -- 0.6 0.4 -- -- B  1.5    2.7     1  F  14.9    5     5.3  Cl  6.4    4.1     2.6  Sc 0.7 0.7 0.7 0.7 0.8 0.1 0.2 3.1 3 0.8 1.1 0.7 V 0.1 0.2 -- 0.2 0.2 0.1 0.2 0.6 1 0.4 0.3 0.2 Cr 0.6 0.2 0.6 0.6 0.6 0.1 0.4 0.4 0.9 0.4 0.5 0.7 Mn 0.2 0.1 0.3 0.2 0.3 0.2 0.2 0.4 0.3 0.3 0.2 0.2 Fe 0.2 0.2 0.7 0.4 0.7 0.1 0.5 0.8 1 0.7 0.7 0.4 Co 2.3 0.9 21.4 12.5 5 0.7 3 80 2.6 1.5 1.2 -- Ni 0.7 0.2 1.1 0.5 0.2 0.1 0.6 0.7 0.1 0.3 0.1 -- Cu 0 0.1 0.2 0.2 0.4 0 -- 0.2 0.1 0.1 0.1 0.2 Zn 0 0 0.1 -- 0.1 -- 0.3 0.1 0.5 0.2 0.1 -- Ga 0.7 0.6 0.3 0.3 0.2 0.4 0.9 0.4 0.7 0.6 0.4 0.3 175  4.8 Figures  Figure 4.1: Location of the Markham Bay spinel localities (1A and 1B). Geology after Blackadar (1967) and Butler (2007). Only the southeastern part of “Spinel Island” was inspected in detail in the present study. Dotted lines represent the extent of exposure at low tide.    Figure 4.2: Spinel-bearing diopsidite band (ca. 2.5 m thick) seen looking West on Spinel Island, north of MacDonald Island, Markham Bay.   176     Figure 4.3: (A: top left) Main area of the diopsidite band at Spinel Island. Note the uneven distribution of phlogopite (Phl)-rich portions. Spinel (Spl) crystals occur in the lower part of the outcrop, where the diopsidite is poorer in phlogopite. Pale-colored calcite (Cal), diopside (Di), and phlogopite occur on the hanging wall of the diopsidite, where it probably grades into marble. (B: top right) Dark blue, euhedral spinel crystals up to 4 cm in size in white calcite, grey diopside, and brown phlogopite. (C: bottom) Spinel (Spl) partly replaced by corundum (Crn) and clinochlore (Chl). Trace gahnite (Gah). Lower part of image contains retrograde minerals margarite (Mrg), zoisite or clinozoisite (Zs/Czs), and tremolite (Tr). Spinel Island, Markham Bay. Backscattered electron image.  177   Figure 4.4: Yellowish-white amphibole-bearing calcite veins and dark amphibole veinlets cross-cutting diopsidite. Unnamed island, Markham Bay.   Figure 4.5: Location of spinel occurrences on Glencoe Island: (2A) Main occurrence, (2B) Marble beach, (2C) Contact, and (2D) Marble Gulch. All spinel occurrences are hosted in Lake Harbour Group (LHG) marbles and calc-silicate rocks. Modified after St-Onge et al. (1999b).  178    Figure 4.6: (A) Main spinel occurrence at Glencoe Island in a 2-meter thick band of marble. (B) Calc-silicate pod (disjointed boudins) at the Glencoe Main occurrence containing spinel (Spl), diopside (Di), calcite (Cal), and phlogopite (Phl) hosted in marble matrix.  179   Figure 4.7: Embayed grain boundaries between forsterite (Fo), pargasite (Prg), and diopside (Di) with apatite (Ap) and retrograde serpentine (Serp). Marble gulch occurrence, Glencoe Island.  180   Figure 4.8: Location of spinel occurrences near Kimmirut sampled for this study (refer to Table 4.1 for localities). Modified after St-Onge et al. (1999a). 181    Figure 4.9: Spinel in dolomitic marble. Soper River spinel occurrence, Kimmirut area.  182   Figure 4.10: Spinel-bearing humitite pods (Hu) in weathered marble outcrop. Soper Falls spinel occurrence, Kimmirut area.      183     Figure 4.11: (A: top left) Spinel (Spl) and white carbonate in humite (Hu). (B: top right) Textural relationship between pargasite (Prg) and humite (Hu) suggesting replacement of the latter by the former. Sample contains talc and magnesite (Mgs). Backscattered electron image. (C: bottom) Dolomite (Dol) bordering magnesite (Mgs) and humite (Hu). Spinel (Spl) is surrounded by a thin rim of clinochlore (Chl). Soper Falls spinel occurrence, Kimmirut area.  184    Figure 4.12: (A) Phlogopite (Phl) porphyroblast in coarse-grained forsterite (Fo) and carbonate rock in the rock face at the Soper Lake mica mine, Kimmirut area. (B) Purple spinel (Spl) with phlogopite (Phl) and forsterite (Fo) in the Soper Lake mica pit, Kimmirut area.  185    Figure 4.13: (A: top) Cobalt-blue spinel outcrop at the Qila occurrence: [3E-1] Calc-silicate pod; [3E-2] Pargasite-calcite rock; [3E-3] Marble. (B: bottom) Vivid blue spinel with white carbonate in calc-silicate rock composed of green pargasite with subordinate greyish scapolite. Qila occurrence [rock unit 3E-1], Kimmirut area. 186    Figure 4.14: Forsterite (Fo) with a diopside (Di) interior corona and partial pargasite (Prg) exterior corona in calcite (Cal) and dolomite (Dol) marble. Trace apatite (Ap). Backscattered electron image. Qila cobalt-spinel marble (sample 3E-3-A), Kimmirut area.    187   Figure 4.15: Silicate-rich spinel-bearing rock (e.g., sample 3F-1), the predominant spinel-bearing unit at the Trailside occurrence, Kimmirut area. It is composed of fine-grained muscovite (Ms) pseudomorphs after an unknown mineral, coarse-grained phlogopite (Phl), calcite (Cal), and spinel (Spl).  188   Figure 4.16: Spinel (Spl) partly altered to corundum (Crn) and clinochlore (Chl). Phlogopite (Phl) has rims of clinochlore. With accessory pyrite (Py) and hosted in calcite (Cal). Backscattered electron image. Trailside occurrence (Zone 3F-1), Kimmirut area.   189   Figure 4.17: Cobalt-blue spinel gemstone, 0.16 carats (0.032 grams) from the Trailside occurrence, Kimmirut area. B.S. Wilson photo. After Wilson (2014).   Figure 4.18: Foliated rock composed of pale yellow calcite, pale brown phlogopite, and cobalt-blue spinel (sample 3F-2) found as float above the main outcrop. Trailside occurrence, Kimmirut area.  190   Figure 4.19: Apparent replacement of forsterite (Fo) by diopside (Di). Backscattered electron image. Sample 3F-M1 from the vicinity of the Trailside occurrence, Kimmirut area.          191   Figure 4.20: T-X(CO2) pseudo-section for sample 2A-SPL-2, diopside-calcite-phlogopite-spinel rock from the Glencoe Main occurrence, Glencoe Island. Mineral assemblage “1” consists of forsterite (Fo), diopside (Di), spinel (Spl), phlogopite (Phl), calcite (Cal) and dolomite (Dol). Grey shading indicates conditions at which spinel is stable. M1A represents peak metamorphic conditions after St-Onge et al. (2007).    192   Figure 4.21: T-X(CO2) pseudo-section for sample 3A-1, representative marble from the Soper River spinel occurrence near Kimmirut containing calcite (Cal) and dolomite (Dol) with subordinate spinel (Spl), pargasite (Prg), phlogopite (Phl), and fine-grained replacements of diopside (Di) and dolomite after an unknown mineral, probably forsterite (Fo). Mineral assemblage “1” consists of Fo, Prg, Spl, Phl, Asp, Cal and Dol. Aspidolite (Asp; Na-end-member of phlogopite) is predicted to occur, but the solution model for phlogopite/biotite does not account for substitution of Na for K. Grey shading indicates conditions at which spinel is stable. M1A represents peak metamorphic conditions after St-Onge et al. (2007).   193   Figure 4.22: T-X(CO2) pseudo-section for sample 3E-3-A, spinel-bearing marble from the Qila occurrence near Kimmirut. Mineral assemblage “1” consists of forsterite (Fo), pargasite (Prg), diopside (Di), spinel (Spl), phlogopite (Phl), calcite (Cal) and dolomite (Dol). Grey shading indicates conditions at which spinel is stable. M1A represents peak metamorphic conditions after St-Onge et al. (2007).    194    Figure 4.23: T-X(CO2) pseudo-section for sample 3F-2, phlogopite-carbonate-spinel rock from the Trailside occurrence near Kimmirut. Mineral assemblage “1” consists of scapolite (Scp), spinel (Spl), phlogopite (Phl), nepheline (Ne), calcite (Cal), and dolomite (Dol). Assemblage “2” consists of Scp, Spl, Phl, Cal, and Dol. Also plotted are geikielite (gk), pargasite (Prg), and corundum (Crn) bearing assemblages. Aspidolite (Asp; Na-end-member of phlogopite) is predicted to occur, but the solution model for phlogopite/biotite does not account for substitution of Na for K. Grey shading indicates conditions at which spinel is stable. M1A represents peak metamorphic conditions after St-Onge et al. (2007).   195   Figure 4.24: Aluminium and silicon content of LHG metasedimentary rocks and reference sedimentary rocks (kaolinite-rich claystone #AR-33, claystone and siltstone averages after López et al. 2005; average shale/clay after Parker 1967; average sandstone after Turekian & Wedepohl 1961). Dashed lines represent Al/Si for the reference sediment values used in protolith composition estimate calculations.    196   Figure 4.25: Ternary plot of the estimated protolith sedimentary composition (wt. %) for LHG spinel-bearing rocks, calc-silicate rocks, marble, Beluga sapphire calc-silicate rock, and Soper River lapis lazuli. For lapis lazuli, the evaporite component is included with carbonate.  197   Figure 4.26: CaO and MgO content of LHG calc-silicate rocks, marbles, Beluga sapphire-bearing calc-silicate rock, and Soper River lapis lazuli. 198    Figure 4.27: (A) Mg-Al-Si ternary diagram of LHG metacarbonate rocks and psammite (relative mol. %); (B) Mg-Al-K ternary diagram of LHG metacarbonate rocks and psammite (relative mol. %). 199    Figure 4.28: Whole rock Co/Fe plotted against Ni/Fe to illustrate the range of Qila and Trailside compared to other Lake Harbour Group metasediments (mostly calc-silicate metacarbonate and marble) and an example set of siliciclastic sedimentary rocks (López et al. 2005). LHG samples with Co or Ni below detection limit are excluded from the diagram.  200   Figure 4.29: Whole rock concentrations of cobalt and iron illustrating the high level of Co enrichment relative to Fe in rocks from Qila and Trailside in comparison to other Lake Harbour Group metasediments analyzed in the present study. Samples with Co below detection limit are excluded from the diagram.  201   Figure 4.30: Whole rock Co/Al vs. Co/Fe in Lake Harbour Group metasediments illustrating the wide range in Co/Al at Qila and Trailside due to low Al contents in Qila Co-rich marble and Trailside Co-rich diopsidite. Sample 3D-2 is excluded due to significant under-representation of Al in the sample relative to the rock unit.  202   Figure 4.31: Iron content of spinel compared to its host rock. Samples insufficiently mineralogically representative of rock unit composition were excluded. Only the spinel-rock pair from Soper Falls humitite deviates from the trend (excluded from the linear regression trendline).     203  Chapter 5. Soper River lapis lazuli: Protolith and effect on quality 5.1 Chapter introduction Bands of lapis lazuli described by Hogarth (1971) and Hogarth & Griffin (1978) occur in the same marble exposure as the Soper River spinel locality north of Kimmirut (Fig. 4.8). Hogarth & Griffin (1978) convincingly demonstrated that the lapis lazuli is a metamorphosed evaporite based on its well-developed layering parallel to regional foliation, the dearth of intrusive rocks in the area, and the elevated concentration of elements typically associated with evaporites. This chapter refines their interpretations of the lapis lazuli protolith and examines how the protolith composition has affected the suitability of Soper River lapis lazuli for use as a gem material. 5.2 Results 5.2.1 Mineralogical composition Lapis lazuli rock samples from the main Soper River lapis lazuli occurrence (Fig. 5.1) were observed with a standard binocular microscope. The rock samples are medium-grained (Fig. 5.2) and dominantly composed of blue haüyne, pale grey diopside, and pale yellowish calcite with subordinate phlogopite and pyrite, consistent with the previous petrographic descriptions of Hogarth (1971).  5.2.2 Whole rock major element composition The major element composition of Soper River lapis lazuli (Table 4.14B) is relatively similar to that of other calc-silicate rocks sampled in the Lake Harbour Group (see Chapters 3 and 4) with the exception of slightly low K concentrations (~ 0.5 wt. % K2O) and exceptionally high Na concentrations (6.33 – 6.96 wt. % Na2O vs ≤ 2.71 wt. % Na2O in other sampled LHG calc-silicate rocks). The high Na concentrations reflect the high abundance of haüyne, a sodium-204  rich feldspathoid. Soper River lapis lazuli contains 0.98 to 1.27 wt. % S, which is primarily incorporated in pyrite and haüyne, and is comparable in concentration only with sulphide-rich metamorphic rocks in the LHG (sulphide-rich diopsidite, psammite; see Chapter 4). 5.2.3 Whole rock trace element composition Compared to other LHG marbles and calc-silicate rocks, the trace element composition of Soper River lapis lazuli rock (Table 4.15B) is characterized by high concentrations of As-Se-Mo (presumably due to the presence of pyrite), Cl, F, an average amount of B (17 µg/g), and very low Co-Ni-Cu. Relatively high sulphur contents (~ 1 wt. %) are due to the presence of haüyne (sulphate-bearing feldspathoid) and pyrite. 5.3 Discussion 5.3.1 Evaporitic origin and protolith Previous work by Hogarth & Griffin (1978) demonstrated that Soper River lapis lazuli is a metamorphosed evaporite layer in marble (see Section 1.6.3). Indeed, the Na2O concentrations in whole rock samples (~ 6 wt. %) are anomalously high; more than twice that in the LHG calc-silicate rock that is the second richest in Na2O (Beluga sapphire-bearing calc-silicate, 2.71 wt. %); as is the concentration of chlorine, 3170 µg/g, almost thrice that of the Cl-bearing pargasite-rich Qila calc-silicate rock (1160 µg/g). Sodium and Cl could have been sourced from a halite component in the original lapis lazuli protolith, but could also be the result of evaporite-derived Na- and Cl-enriched fluid. The high abundance of haüyne (the Na-rich mineral phase) in the lapis lazuli bands over a large geographical area is more consistent with a halite-bearing protolith. Under this assumption, an estimation of the original protolith composition (method in Chapter 4; Table 4.16B; Fig. 4.25) is an evaporitic dolomitic marl, composed of ca. (by weight) 40% carbonate (dolomite ≈ calcite to dolomite > calcite), 10 % halite, and 50 % siliciclastic mud. 205  Hogarth (1971) proposed the presence of anhydrite in the protolith to explain the relatively high S concentrations (~1 wt. %), an element that is a fundamental component of haüyne. Therefore, lapis lazuli rock (and therefore the protolith) is relatively similar in major element composition (i.e., characteristic of metamorphosed dolomitic marl) to spinel-bearing calc-silicates (Chapter 4) and Beluga sapphire-bearing calc-silicate rock (Chapter 3), but differs from the spinel-bearing rocks in lower Mg/Ca and Al/Si, and differs from both the spinel- and sapphire-bearing rocks in its high Na and S content. The abundance of Na and S, in this case interpreted to be due to the presence of halite and anhydrite evaporites in the dolomitic marl protolith, is key to forming lapis lazuli. 5.3.2 Composition of protolith vs. gem quality of lapis lazuli Soper River lapis lazuli is not suitable as a commercial gem material due to its low haüyne concentration (i.e., Fig. 5.2) relative to that of significantly higher quality materials from producing deposits (e.g., Sar-e-Sang, Afghanistan). The small crystal size of haüyne at Soper River precludes the use of single crystals as gem materials. The high percentage of calcite and diopside are, therefore, detrimental to lapis lazuli gem quality, and the abundance of these minerals is interpreted to reflect the composition of the protolith (carbonate content, carbonate species, character of the siliciclastic component such as relative abundance of Al and Si).  Lapis lazuli formed by metasomatism at two of the three major deposits (Chile, Coenraads & Canut de Bon 2000; Russia, Deer et al. 2004 and references therein) and metasomatism possibly occurred at Sar-e-Sang, Afghanistan, especially for haüyne formed at the contact with granite/pegmatite (Faryad 2002 and references therein). While haüyne can occur in metasediments (sensu stricto), the protolith composition of lapis lazuli at Soper River is unsuitable for the genesis of gem grade lapis lazuli. Haüyne contains an equal molar abundance 206  of Al and Si (Al/Si = 1), but siliciclastic sediments generally have lower Al/Si (kaolinite-rich claystone Al/Si = 0.78 mol/mol, López et al. 2005; average shale/clay Al/Si = 0.46, Parker 1967). Two samples of Soper River lapis lazuli have Al/Si ≈ 0.32. Therefore, higher concentrations of haüyne in lapis lazuli rock would be favoured by higher proportions of Al relative to Si (i.e., higher shale/clay siliciclastic component relative to silt/sand in protolith), an abundance of evaporites (i.e., halite and anhydrite, providing the key elements Na and S, respectively, required for haüyne formation), and a lower proportion of carbonate. Metasedimentary (sensu stricto) gem-grade lapis lazuli is likely possible, but non-ideal proportions of Al/Si (siliciclastic component), carbonates, and potentially evaporites in the Soper River lapis lazuli protolith are expected to have prevented the genesis of high-quality gem lapis lazuli. 5.4 Conclusions Soper River lapis lazuli is interpreted to consist of a metamorphosed dolomite-bearing to dolomitic marl that contained a significant evaporitic component (halite and subordinate anhydrite). The enrichment of Na and S from evaporite is required to form haüyne, the blue mineral that is the fundamental component of lapis lazuli rock. The Soper River lapis lazuli deposit is of relatively low gem quality due to its relatively low concentration of haüyne compared to commercial grade gem-quality lapis lazuli from other localities. Low Al/Si (interpreted to be a result of a silty siliciclastic component) and high carbonate content in the protolith prevented the genesis of high-quality gem lapis lazuli at Soper River.  207  5.5 Figures  Figure 5.1: Part of the weathered lapis lazuli band at the main Soper River occurrence, looking SE toward the violet spinel occurrence.   Figure 5.2: Lapis lazuli rock from Soper River. The rock is composed of pale yellowish calcite, grey diopside, and blue haüyne. 208  Chapter 6. General deposit model and exploration criteria  6.1 Fundamental assumptions about LHG metacarbonate protoliths The Kimmirut calc-silicate-hosted sapphire occurrences and most spinel occurrences are interpreted to be metasedimentary on the basis of their isolation within a metacarbonate sequence with no proximal magmatic intrusions or textural/mineralogical evidence for metasomatism (Chapters 3 and 4). Most of these occurrences are sufficiently homogeneous (or show distinct separations between calc-silicates and marble) to warrant the assumption that their whole rock major element composition is representative of the original protolith composition, although the possibility of diagenetic and metamorphic transformations (especially with regards to more mobile elements) should not be discounted. It should be noted that deformation could also spatially redistribute rock-forming components, such that mineralogical/major element compositional zoning at some occurrences (e.g., Trailside spinel occurrence, Chapter 4) may not be representative of the original protolith. However, hypothetical protolith compositions estimated from the metasedimentary rock bulk compositions would produce identical mineral assemblages if subjected to the same metamorphic conditions. Therefore, we consider the assumption that whole rock major-element compositions are representative of protolithic compositions, in most cases, to be a reasonable and useful new approach to understanding metasedimentary gem deposits.  Hogarth & Griffin (1978) supported the probable metasedimentary and evaporitic origin of Soper River lapis lazuli by citing the following evidence: (1) layering is well-developed and parallel to the regional foliation, suggesting that it is metasedimentary; (2) no intrusive rocks occur in the area, which argues against a contact metasomatic origin; and (3) the abundances of Na, K, S, Cl, Br, F, and Fe are consistent with evaporite-related sediments.  209   Using the aforementioned assumptions, estimates of protolith compositions were calculated for the following LHG rock types (Chapter 4): (1) marble and calc-silicate rocks barren of gem minerals; (2) spinel-bearing marble and calc-silicate rocks (including humite- and forsterite-rich metacarbonates); (3) sapphire-bearing calc-silicate rocks; and (3) lapis lazuli. Estimates were calculated by estimating the character of the siliciclastic component by comparing sample Al/Si to that of sedimentary averages and re-calculating the expected original concentration of carbonates (calcite, dolomite, magnesite) and, where applicable, halite. 6.2 Gem deposit model 6.2.1 Sedimentary protoliths and metasomatic occurrences Corundum, spinel, and lapis lazuli occurrences in the LHG are all related to metasedimentary carbonates. More specifically, most occurrences are interpreted to have the following protoliths: (A) impure dolomite-bearing limestone to dolostone; (B) dolomitic marl; (C) magnesite-rich evaporitic marl; and (D) evaporite (halite and anhydrite)-bearing dolomite-rich marl (Fig. 6.1). Spinel at “Spinel Island” in Markham Bay formed from contact metasomatism of marble and syenogranite, which is probably related to the Cumberland batholith and, together with the metasomatic black spinel occurrence on a nearby unnamed island, are the most significant metasomatic spinel occurrences reported from the LHG, although the exact origin of the Soper River mica mine and Waddell Bay occurrences could not be assessed. The Contact occurrence on Glencoe Island is interpreted to have formed from the bimetasomatic reaction of contiguous gneiss and marble (Chapter 4). In metasedimentary occurrences, metacarbonates containing gem mineral species have carbonate-dominant protoliths (spinel) or mixed mud-carbonate(-evaporite) protoliths (corundum, spinel, lapis lazuli; Fig. 4.25). 210  6.2.2 Geochemical controls on gem mineral genesis  The gem minerals corundum, spinel, and haüyne (a feldspathoid) occur in silica undersaturated environments. Corundum near Kimmirut occurs in localized alteration zones within a calc-silicate rock, which formed during retrograde metamorphism. It should be noted that silica undersaturation in the whole rock at Kimmirut, during peak granulite facies metamorphism, was a key factor in forming the precursor assemblage containing nepheline, a key reactant in gem corundum genesis at the locality (Chapter 3). Therefore, silica undersaturation is a common feature of all LHG corundum-, spinel-, and haüyne-bearing rocks. Silica undersaturation in LHG gem-mineral-bearing rocks is interpreted to have been achieved in three different ways (Chapters 3-5): (1) The mixing of siliciclastic and carbonate sediments in the protolith (i.e., marl) with a high mud content, giving the rock a high alumina to silica ratio (e.g., Kimmirut sapphire, metasedimentary spinel-bearing calc-silicates, lapis lazuli); (2) the occurrence of small amounts of siliciclastic impurities in a limestone or dolostone protolith (e.g., spinel-bearing marbles); and (3)  the bimetasomatic reaction of an Al-bearing, silica-rich rock with a contiguous marble (e.g., syenogranite-marble on Spinel Island, gneiss-marble at the Glencoe Contact occurrence). In each case, decarbonation reactions led to the formation of various Ca- and Mg-silicate minerals and produced the Si-undersaturated conditions necessary for gem formation (oxide phase, spinel; feldspathoid, haüyne), or for generating a suitable precursor to corundum-bearing rock (nepheline-bearing calc-silicate rock). A low potassium abundance (K/Al < 1 mol/mol) favors the formation of spinel, whereas high whole rock K/Al is hypothesized to result in the preferential incorporation of Al into phlogopite. The difference in K/Al between spinel-bearing and spinel-absent marbles is particularly significant in the LHG, likely due to the higher concentration of Si relative to Al 211  compared to the silicate-rich metacarbonates, in which Al/Si is the dominant differentiator between spinel-bearing and spinel-absent rocks (K/Al may also be a control on spinel concentration in calc-silicates; Chapter 4).  However, in the case of bimetasomatic spinel- and phlogopite-bearing diopsidite on Spinel Island, Mg (i.e., dolomite in the marble) rather than K is interpreted to have been the limiting element in the formation of spinel and additional diopside rather than phlogopite. XCO2 does not affect the presence of spinel in pseudo-sections of two silicate-rich metacarbonates, but it is an important control on the presence or absence of spinel in marbles at peak metamorphic conditions (Chapter 4). In lapis lazuli, the contribution of Na and sulphate by evaporites, i.e., halite and anhydrite, is required to form haüyne (Hogarth & Griffin 1978, Chapter 5). The Na content of Soper River lapis lazuli (~ 6 wt. % Na2O) is considerably higher than that of other calc-silicate rocks in the LHG (≤ 2.71 wt. % Na2O; Chapter 5). Soper River lapis lazuli is unsuitable for use as a commercial gemstone due to its relatively high amount of calcite and diopside impurities; a protolith with a higher Al/Si fraction and lower carbonate content would be more favorable to gem lapis lazuli genesis. 6.2.3 Metamorphic controls on gem mineral genesis  In a broad sense, two conditions are required to form a metamorphic gem mineral occurrence. Firstly, a rock (or multiple rocks in the case of metasomatic occurrences) must be of favorable geochemistry and subjected to specific pressure and temperature in order to form the gem-bearing assemblage. Secondly, the gem mineral must not be significantly destroyed by subsequent metamorphic processes (e.g., fluid influx during retrograde path at P-T-x where the mineral is not stable, or loss of gem quality from deformation of transparent gem materials). Calc-silicate hosted sapphires formed through a multi-step process requiring two retrograde 212  mineralization events at specific temperatures, while haüyne and metamorphic (sensu stricto) spinels are interpreted to have formed at granulite peak metamorphic conditions. None of the gem mineral species in the LHG studied thus far show significant levels of retrograde alteration; however spinel at a number of occurrences is heavily fractured.   Near Kimmirut, three sequential metamorphic reactions led to the formation of gem-quality corundum (Chapter 3). Granulite facies (peak) metamorphism of the dolomitic marl formed an assemblage of nepheline, diopside, and K-feldspar (inferred). This assemblage was partly retrogressed to phlogopite, oligoclase, calcite, and scapolite as a result of CO2-, H2O-, Cl-, and F-bearing fluid influx. The zircon U-Pb age from this first retrograde assemblage indicates crystallization ~30 Ma after the end of M2 metamorphism (of St-Onge et al. 2007), and therefore at P < 6.2 kbar and T < 720 °C. The formation of scapolite in contact with nepheline is a necessary precursor to the second retrograde reaction, where corundum is formed: Scapolite (Me57) + 0.84 nepheline + 0.60 H2O + 2.02 CO2  0.76 corundum + 2.35 calcite + 2.14 albite + 0.60 muscovite + 0.66 Cl The corundum-forming reaction occurs at lower P-T than the first retrograde assemblage. A thermodynamic model of the break-down of scapolite and nepheline to corundum, calcite, and albite (with scapolite modeled as end-member meionite) indicates that corundum can only form in a < 100 °C temperature window, bound by scapolite-nepheline stability at higher temperature and the formation of Al-silicate at lower temperature (Chapter 3).  Spinel studied at just over a dozen occurrences in the LHG (Chapter 4) is interpreted to have formed during granulite facies peak metamorphism based on stable mineral assemblages and, in several cases, thermodynamic modelling. Some occurrences (e.g., Spinel Island) are assumed to have formed under these conditions based on the stability of similar assemblages that 213  were thermodynamically modeled (Glencoe Island main occurrence). The timing and P-T conditions of the Markham Bay unnamed island occurrence and the Waddell Bay occurrence could not be ascertained. Several, but not all, spinel-bearing assemblages in the Kimmirut area have peak metamorphic forsterite with subsequent (peak metamorphic or retrograde) diopside and/or pargasite, either as reaction rims or as well-developed crystals. Spinel at Spinel Island and the Trailside occurrence is locally, and in minor part, replaced by retrograde corundum in association with other minerals. The corundum alteration of spinel is not so significant as to affect the quantity or quality (i.e., lack of impurities) of spinel at these occurrences, and there is no indication that the LHG has potential for gem corundum deposits formed from the retrograde break-down of spinel.  Haüyne / lapis lazuli at Soper River formed at or close to peak metamorphic conditions (granulite facies; Hogarth & Griffin 1978). 6.2.4 Controls on chromophore concentrations Gem-quality corundum near Kimmirut varies from colourless to blue (Fe-Ti chromophore, Fritsch & Rossman 1988), commonly as oscillatory and sector zoning within single crystals but also as uniquely blue or colourless crystals, and yellow (Fe chromophore, Fritsch & Rossman 1988; Fig. 3.2). Iron and Ti activity varied during corundum crystallization, as evidenced by oscillatory zoning, however the cause of this variation is not known. Rare, pale greyish pink corundum found during exploration (True North Gems 2007) presumably contains small amounts of chromium (the chromophore in pink and red corundum, Fritsch & Rossman 1987). However, the dominant Cr-bearing phases (phlogopite and diopside) in Kimmirut corundum-bearing calc-silicate rocks are not part of the corundum-forming reaction; since corundum formed from the break-down of Cr-poor phases (scapolite, nepheline), the potential 214  for pink gem corundum (i.e., pink sapphire) in these rocks is low. The whole rock Cr concentration in the Beluga sapphire occurrence is relatively high for a marl protolith (average 108 µg/g), but Cr/Al (0.00075 mol/mol) is substantially lower than that of ruby (red gem corundum) bearing rocks (e.g., ruby mica schist Cr/Al = 0.002-0.006, Yakymchuk & Szilas 2018; ruby pargasite schist Cr/Al = 0.050, Wang et al. 2017).  Spinel in the LHG is blue, violet, purple, black, and rarely, cobalt-blue in colour.  With increasing Fe content, spinel varies from colourless to pale lilac, sky blue, green, deep green and black (Kleišmantas & Daukšytė 2016, Hålenius et al. 2002; black spinel e.g., Van Velthuizen 1993), and colour can vary with Fe coordination, charge, or interplay between Fe2+ and Fe3+ (D’Ippolito et al. 2015). In general, however, relatively high Fe concentrations in spinel do not result in attractive colours for use in the gemstone industry. Spinels from the LHG contain roughly four times the iron concentration of their host rocks, with some exceptions, and Chapter 4 discussed the importance of lower than expected Fe concentrations in spinel-bearing rocks relative to the Fe concentrations in a typical shale/carbonate protolith in order to favour spinel gem quality.  Vivid blue spinel at two occurrences near Kimmirut - Trailside and Qila - are cobalt-enriched (up to 0.07 wt. % CoO), which is responsible for its intense colouration (see D’Ippolito et al. 2015, Chauviré et al. 2015). The bright blue colour of spinel at these occurrences makes them of significant gemological and economic interest. Chapter 4 identified a localized enrichment of both Co and Ni at Qila and Trailside relative to other LHG metacarbonates (ranges of 9 to 29 µg/g vs. <1 to 6 µg/g cobalt; 10 to 47 µg/g vs. < 1 to 23 µg/g nickel, respectively), with limited diffusion of these trace elements (≤ 1 m scale) during metamorphism. The cobalt and nickel enrichment in these metacarbonates, which are isolated within a marble sequence, are 215  interpreted to be representative of the protolith composition. Concentrations of Co (especially) and Ni at Qila and Trailside are anomalously high for metacarbonate protoliths, at least double the estimates for similar protoliths of average composition, and the rocks are relatively impoverished in Fe, Mn, V, Cr, and Cu. The enrichment of Co and Ni, and depletion of V and Cr can take place under oxic early diagenetic conditions in the protolith sediment, the favoured hypothesis, but this is difficult to reconcile with a lack of coeval Mn enrichment, which is expected to occur under the same conditions (Chapter 4). 6.3 Exploration criteria & methods 6.3.1 Kimmirut-type sapphire deposits Potential for Kimmirut-type sapphire deposits is strongly restricted by the P-T history of a metacarbonate-bearing terrane. On Baffin Island, potential for such deposits are expected to be proximal to the thrust fault separating the Lake Harbour Group and Narsajuaq terranes, where the retrograde amphibolite facies overprint of peak metamorphic assemblages is most pervasive. Known deposits have a relatively small footprint (calc-silicate pods up to 4.2 × 3.7 m), and occur throughout the True North Gems claim property (44 corundum showings discovered to date). Scapolite, a rock-forming constituent of calc-silicate pods in the area (of which approximately 0.5% contain corundum mineralization visible in outcrop), fluoresces bright yellow in long wave ultraviolet light. These fluorescing properties were used by True North Gems to explore for sapphire-bearing calc-silicate pods by prospecting the area with UV lamps in low light conditions (Lepage & Davison 2007). Interestingly, an occurrence of purplish-pink corundum in New York state, which also formed from scapolite-nepheline break-down, occurs in fluorescent scapolite (see Chapter 3). Ultraviolet light exploration has proven to be a valuable prospecting tool, however corundum could theoretically be found associated with non-fluorescent scapolite. 216  In normal light, the calc-silicate rocks have a distinctive mottled appearance caused by the contrast between darker minerals (purplish-brown diopside, brown phlogopite) and lighter ones (scapolite, calcite, albite-rich corundum-bearing zones). Turner et al. (2017) successfully mapped distinct phlogopite, muscovite, scapolite, prehnite, and zeolite hyperspectral data domains using the Spectral Angle Mapper algorithm. Prehnite and zeolites represent low temperature alteration products closely associated with the corundum-bearing zone. Turner et al. (2017) suggested that these results could be used to conduct regional spectroscopic imaging in order to explore for Kimmirut-type gem corundum deposits in southern Baffin Island. 6.3.2 Gem spinel deposits Spinel in the LHG occurs in two types of rock: (1) silicate-rich Mg-bearing metacarbonates with high Al/Si (rock forming minerals i.e., diopside, pargasite, forsterite, phlogopite, humite, calcite, spinel); and (2) impure dolomite-bearing or dolomitic marbles (calcite/dolomite with subordinate silicates such as forsterite, pargasite, diopside, and phlogopite) with K/Al << 1. These rocks most commonly occur as pods within marble sequences, while some spinel-bearing diopsidites occur at the contact of marble and a silicate-rich rock (e.g., syenogranite or gneiss) or within a calc-silicate rock (i.e., metasomatic occurrence on unnamed island in Markham Bay). Magnesian marble sequences, especially where included by abundant calc-silicate pods (or other metamorphosed Mg-rich marls such as humitite or forsterite-rich rock), are expected to be the most prospective for spinel. Prospective rock units may be clustered together as a result of favourable geochemical compositions in the protolith stratigraphy. Local variations of K/Al in phlogopite-rich layers may lead to spinel formation. The highly localized nature of cobalt enrichment at cobalt-blue spinel occurrences make exploration 217  difficult, however the occurrence of two unique cobalt-enriched metacarbonates within 2 km is noteworthy. Whole rock geochemical sampling on a regional scale would be cost prohibitive for gemstone exploration, but regional geology surveys may happen to identify Co-enriched metacarbonates – good potential targets for gem spinel exploration. 6.3.3 Lapis lazuli deposits Soper River lapis lazuli is interpreted to be metamorphosed evaporite-bearing dolomitic marl but this does not provide helpful exploration criteria because the whole rock composition of the rock at Soper River is unfavourable for its use in the commercial gemstone industry (i.e., abundant diopside and calcite impurities). Lapis lazuli, relative to gem-quality sapphire and spinel, has a very low value per unit of volume. Therefore, larger volumes of high quality lapis lazuli are required to form a deposit. Considering the excellent level of rock exposure on southern Baffin Island, identification of bright blue rock within marble is currently the only viable exploration method. 6.3.4 Exploration using aerial surveys Southern Baffin Island is an extremely remote region with limited infrastructure and a short (~two month) field season. The harsh climate offers one significant advantage: the rock is extraordinarily well-exposed and is scarcely covered by lichens, plants, or soil. This makes the area uniquely suited to exploration using aerial surveys. Harris et al. (2010) used airborne hyperspectral data to produce spectral maps identifying calcite-, dolomite-, and diopside-rich domains within a carbonate sequence in the LHG on southern Baffin Island. Turner et al. (2017) proposed the use of this method in combination with hyperspectral signatures of scapolite, phlogopite, muscovite, and other minerals to explore for gem corundum. These methods could be used in aerial surveys over marble-rich regions of Baffin Island to efficiently and thoroughly 218  seek out gem corundum and spinel exploration targets: magnesian marble sequences with abundant calc-silicate units. High resolution colour imagery could be collected simultaneously and be used to explore for blue lapis lazuli layers in marble.  Local-scale aerial surveys could also aid in exploration. The author used a DJITM Mavic Pro unmanned aerial vehicle (UAV) video/photo camera to successfully explore for gem-quality olivine in the mountainous terrain of British Columbia, Canada. In good conditions, it was flown over 3 km away, 300 meters in elevation gain, and brought within 3 meters of basalt talus, where it successfully photographed ca. 10 cm wide mantle xenoliths in basalt blocks. Gem peridot exploration using this UAV took four days and saved roughly two weeks of field work when compared to traditional exploration. A portable UAV camera could help in field reconnaissance for calc-silicate units, potentially interesting mineralization, and blue lapis lazuli layers in southern Baffin Island. 6.4 Conclusions Sapphire (gem corundum), spinel, and haüyne occurrences in the LHG are all genetically related to a metamorphosed carbonate sequence. Most occurrences of these gem minerals are uniquely metasedimentary (i.e., formed from the metamorphism of a sedimentary protolith), while a few spinel occurrences formed from metasomatic reactions between Si-Al-rich rock (syenogranite or gneiss) and marble. Metasedimentary corundum, spinel, and haüyne occurrences have relatively similar protoliths: primarily dolomitic marls with a high Al/Si relative abundance (interpreted as sandy mud to clay siliciclastic fraction in the protolith). Kimmirut-type sapphire corundum deposits can only be formed by a multi-step metamorphic process under three different and specific P-T conditions (peak granulite facies assemblage transformed by two stages of subsequent retrograde metamorphism). Lapis lazuli formation at 219  peak or near-peak metamorphic conditions required the presence of evaporites to provide Na and S for the blue mineral haüyne. Spinel formed in Mg-bearing calc-silicates (high Al/Si) and in impure dolomitic marbles with very low K/Al (or subject to low XCO2) during granulite facies peak metamorphism. Potential for Kimmirut-type sapphire deposits is expected to be restricted to marbles proximal to the thrust fault separating the LHG from the Narsajuaq Arc, where retrograde upper amphibolite facies mineralization is most pervasive. Spinel and Kimmirut-type sapphire deposits are expected to likely be found in dolomitic marble sequences rich in calc-silicate layers. The potential occurrence of lapis lazuli is more difficult to predict. Aerial hyperspectral and photographic surveys are well-suited to gemstone exploration in this remote region thanks to excellent rock exposure with minimal sedimentary or plant/lichen cover. Spectral mapping of dolomite-, diopside-, phlogopite-, and scapolite-rich domains in LHG metacarbonate sequences is expected to provide exploration targets. Ground based exploration efforts could be aided by the use of a portable UAV. 220  6.5 Figures  Figure 6.1: The proposed genetic model for Lake Harbour Group gem mineral occurrences. Stages containing gem minerals are represented in blue, and those devoid of gem minerals in black.  221  Chapter 7. Conclusions and future work This dissertation presents new genetic models for Kimmurut-type sapphire (a new type of gem corundum deposit) and gem spinel, and a refined model for the genesis of lapis lazuli (haüyne). These three gem minerals occur in silica undersaturated rocks and generally occur in rocks with similar protoliths: dolomitic marls, transformed into silicate-rich metacarbonates by metamorphism. Spinel also occurs in dolomitic limestone, and a few metasomatic occurrences are known. Variations in the abundances of Na, K, Mg, Al, Si, and S play an important role in determining whether corundum, spinel, haüyne, or no gem minerals will form. The metamorphic P-T-X history of the terrane, including local variations in retrograde fluid influx, also plays a major role in gem genesis in the LHG. Haüyne and spinel formed at peak metamorphic conditions while gem corundum was formed by a three step metamorphic process involving two stages of retrogression. The protolith for Beluga sapphire-bearing calc-silicate rock is interpreted to be dolomitic marl. New petrological and geochemical data indicated that this deposit does not consist of metamorphosed evaporite, as proposed by previous research. A model for the genesis of sapphire within this calc-silicate rock was established using a combination of petrography, thermodynamic modelling and zircon age dating. A peak metamorphic nepheline- and diopside-bearing assemblage was partly replaced by oligoclase-phlogopite-calcite-scapolite as a result of fluid influx shortly following M2 retrograde metamorphism. This partial retrogression placed scapolite and nepheline in direct contact. A subsequent fluid influx, at yet lower metamorphic grade, produced corundum, albite, muscovite, and calcite from the break-down of scapolite and nepheline. Obscuring of the peak metamorphic assemblage by intense retrograde alteration has introduced some possibility for error in connecting the modeled reactions with observed mineral 222  assemblages; notably, K-feldspar is predicted to occur in the peak metamorphic assemblage, but it was not observed in the Beluga and Bowhead samples. Similarly, the formation of corundum from scapolite-nepheline break-down is inferred at Beluga, since no relict nepheline was found. However, new data from a similar locality in New York demonstrated an undeniable textural relation between corundum-albite-calcite formed from breaking down of nepheline and scapolite at their contact. The thrust fault separating the LHG and Narsajuaq terranes appears to be the most prospective for similar deposits, since the retrograde, amphibolite facies overprint of the granulite peak assemblages is most pervasive here. Future research and exploration may uncover new regions of sapphire-bearing potential. Spinel occurrences in the LHG are interpreted to be metasedimentary, with the exception of two metasomatic occurrences (e.g., syenogranite-marble bimetasomatism at Spinel Island). Spinel-bearing metacarbonate rocks are interpreted to have the following sedimentary protoliths: (1) impure dolomite-bearing and dolomitic limestone (marble); (2) dolomitic marl (calc-silicate rocks); and (3) evaporitic magnesitic marl (Soper Falls humitite). In Mg-bearing impure marbles, spinel genesis can be controlled by K activity and XCO2, and is not significantly affected by Al/Si. In Mg-bearing calc-silicates, however, spinel only occurs at high Al/Si; higher than that for sapphire-, haüyne-, or non-gem-bearing metacarbonates. Spinel occurrences at Waddell Bay and Hall Peninsula were not studied in the field; doing so could provide additional, useful petrological data. Spinel is interpreted to have formed during granulite facies peak metamorphism. For multiple occurrences, this is supported by calculated T-X pseudo-sections. However, the very coarse grain size, very heterogeneous nature, and often limited surficial exposure of the spinel occurrences prevented the use of thermodynamic modelling in many cases. Interpretations on the timing of spinel genesis could be reinforced with a significantly 223  expanded sampling program exposing more rock and gathering larger samples, for both thermodynamic modelling and zircon extraction (U-Pb age dating).  This dissertation presents the first whole-rock geochemical dataset for cobalt-enriched spinel-bearing rocks, shedding new light on the possible origins of vivid blue Co-enriched spinel. The highly heterogeneous and localized nature of Co-enrichment in metacarbonates is interpreted to represent early enrichment (protolith, diagenesis, early metamorphism), and contradicts previous research for a similar locality in Vietnam suggesting transport in evaporitic fluids from either within marbles, or from regional Co-Ni-rich rocks. In all cases, the cobalt enrichment is associated with a coeval enrichment in nickel, and relatively low concentrations of Fe, Mn, V, Cr, and Cu. The trace metal composition of these rocks is distinctly different from most typical sources of Co-Ni enrichment, and most resembles (though not perfectly) the effects of oxic early diagenesis. Since the protoliths have undergone diagenetic, structural, and metamorphic transformations, such interpretations are ultimately speculative.   In agreement with previous authors, the exceptionally high Na content and relative S enrichment of haüyne-bearing rock (lapis lazuli) at Soper River supports a meta-evaporitic origin. Low Al/Si and high carbonate content in the protolith prevented the genesis of high-quality gem lapis lazuli, which has significantly higher haüyne concentrations. The present research integrates data from multiple types of gem occurrences and compares them to similar rocks barren of gem minerals within the same metamorphic terrane. This new approach has yielded significant new contributions to the understanding of the metamorphic and geochemical conditions of gem genesis and successfully answered broad questions raised by previous authors (i.e., conditions of spinel genesis in Giuliani et al. 2017). The estimation of protolith composition using whole rock concentrations of relatively immobile 224  elements presents a new approach to the analysis of gem deposits occurring in metamorphosed sedimentary rocks. The analysis of numerous whole rock trace element compositions helped set a reference benchmark for the assessment of unusual Co-enriched, vivid-blue-spinel-bearing rocks. Larger scale application of these methods could help better understand the trace element composition of individual gem minerals within metasedimentary rocks. Excellent opportunities exist for future research. Firstly, an investigation of the distribution of gemstone occurrences within the context of structural geology would contribute important and practical information on the genesis of metamorphic gemstones. For example:  does deformation of silicate-rich layers affect gem potential (e.g., thinning of layers increasing the surface area with carbonates for metamorphic reactions)? Are gem occurrences traceable along stratigraphy or structural domains? To what degree has deformation modified the composition and distribution of formerly continuous layers (i.e., sapphire-bearing calc-silicates)?  The spinel deposit model would benefit from the addition of petrological/geochemical data from actively mined in situ gem spinel deposits. Lastly, the exact origin of cobalt enrichment in metacarbonates remains mysterious. Detailed petrological/geochemical studies on Vietnamese cobalt-blue spinel occurrences, using the methods presented in this dissertation, could shed new light on this phenomenon.  225  References Armstrong, J.T. (1988): Quantitative analysis of silicates and oxide minerals: Comparison of Monte-Carlo, ZAF and Phi-Rho-Z procedures. In: Newbury, D.E. Microbeam Analysis, p. 239-246. Baccar, M.B., Fritz, B., & Madé, B. (1993): Diagenetic albitization of K-feldspar and plagioclase in sandstone reservoirs: Thermodynamic and kinetic modeling. Journal of Sedimentary Petrology 63, 1100-1109. Belley, P.M. & Bourdeau, J. (2014): Blue calcite skarns as potential gem deposits: Examples from the Central Metasedimentary Belt, Grenville Province, Canada. In: Geological Society of America annual meeting, Vancouver, Canada, 19-22 October, 2014, abstract # 165-8. Belley, P.M., Dzikowski, T.J., Fagan, A., Cempirek, J., Groat, L.A., Mortensen, J.K., Fayek, M., Giuliani, G., Fallick, A.E., & Gertzbein, P. (2017): Origin of scapolite-hosted sapphire (corundum) near Kimmirut, Baffin Island, Nunavut, Canada. Canadian Mineralogist 55, 669-699. Belley, P.M., O’Neil, J., & Hattori, K. (2016): Spinel and aspidolite from the Des Cèdres Dam occurrence, Notre-Dame-du-Laus, Québec, Canada. Rocks & Minerals 91, 448-452. Belley, P.M., Grice, J.D., Fayek, M., Kowalski, P.M. & Grew, E.S. (2014): A new occurrence of the borosilicate serendibite in tourmaline-bearing calc-silicate rocks, Portage-du-Fort marble, Grenville Province, Québec: evolution of boron isotope and tourmaline compositions in a metamorphic context. Canadian Mineralogist 52, 595-616. 226  Beranek, L.P. & Mortensen, J.K. (2011): The timing and provenance record of the Late Permian Klondike orogeny in northwestern Canada and arc-continent collision along western North America. Tectonics 30, TC5017, doi:10.1029/2010TC002849. Berman, R.G. (1991): Thermobarometry using multiequilibrium calculations: a new technique with petrologic applications. Canadian Mineralogist 29, 833-855. Berman, R.G. (1988): Internally-consistent thermodynamic data for minerals in the system Na2O-K2O-CaO-MgO-FeO-Fe2O3-Al2O3-SiO2-TiO2-H2O-CO2. Journal of Petrology 29, 445-522. Blackadar, R.G. (1967): Geology of Mingo Lake-MacDonald Island map-area, Baffin Island, District of Franklin. Geological Survey of Canada, Memoir 345, 54 p. Blauwet, D. (2006): The sapphire and spinel deposit of An Phu, Luc Yen Mining District, Yenbai Province, Vietnam. Mineralogical Record 37, 225-2006.  Brandon, M.T., Roden-Tice, M.K., & Garver, J.I. (1998): Late Cenozoic exhumation of the Cascadia accretionary wedge in the Olympic Mountains, northwest Washington State. Geological Society of America Bulletin 110, 985-1009. Butler, J.P. (2007): Petrogenesis of nepheline- and scapolite-bearing metacarbonates from southwestern Baffin Island, Nunavut, Canada. B.Sc. thesis, Dalhousie University, Halifax, Nova Scotia, Canada. Cabral, A.R., Wiedenbeck, M., Rios, F.J., Gomes, A.A.S. Jr, Filho, O.G.R. & Jones, R.D. (2012): Talc mineralisation associated with soft hematite ore, Gongo Soco deposit, Minas Gerais, Brazil: petrography, mineral chemistry and boron-isotope composition of tourmaline. Mineralium Deposita 47, 411-424. 227  Carmichael, R.S. (1989): Physical properties of rocks and minerals. CRC Press, Inc., Boca Raton, Florida, USA. 741 p. Catanzaro, E.J., Champion, C.E., Garner, E.L., Marinenko, G., Sappenfield, K.M., & Shields, W.R. (1970): Boric acid: isotopic and assay standard reference materials. National Bureau of Standards (US) Special Publications, 260-17, 1–71. Chamberlain, S.C., Walter, M., Bailey, D.G., & Robinson G.W. (2015): A new collecting site at the Rose Road Locality, Pitcairn, St. Lawrence County, New York. Rochester Mineralogical Symposium, Rochester, NY. Abstract. Rocks & Minerals 90, 440-446. Chaussidon, M. & Albarède, F. (1992): Secular boron isotope variations in the continental crust: An ion microprobe study. Earth and Planetary Science Letters 108, 229–241. Chauviré, B., Rondeau, B., Fritsch, E., Ressigeac, P., & Devidal, J.-L. (2015): Blue spinel from the Luc Yen District of Vietnam. Gems & Gemology 51, 2-17. Coenraads, R.R. & Canut de Bon, C. (2000): Lapis lazuli from the Coquimbo region, Chile. Gems & Gemology 36, 28-41. Connolly, J.A.D. (2009): The geodynamic equation of state: what and how. Geochemistry, Geophysics, Geosystems 10, Q10014 DOI:10.1029/2009GC002540 Corriveau, L. (2013): Structure of the Central Metasedimentary Belt in Québec, Grenville Province; an example from the analysis of high-grade metamorphic terranes. Geological Survey of Canada Bulletin 586 (in French). 251 p. D’Ippolito, V., Andreozzi, G.B., Hålenius, U., Skogby, H., Hametner, K., & Günther, D. (2015): Colour mechanisms in spinel: cobalt and iron interplay for the blue colour. Physics and Chemistry of Minerals 42, 431-439.  228  Deer, W.A, Howie, R.A., Wise, W.S., & Zussman, J. (2004): Rock-forming minerals Vol. 4B: Framework Silicates, 2nd ed., p. 344-345. Drake, M.J. & Weill, D.F. (1972): New rare earth element standards for electron microprobe analysis. Chemical Geology 10, 179-181. Dzikowski, T.J., Cempirek, J., Groat, L.A., Dipple, G.M., Giuliani, G. (2014): Origin of gem corundum in calcite marble: The Revelstoke occurrence in the Canadian Cordillera of British Columbia. Lithos 198-199, 281-297. Dzikowski, T.J. (2013): A comparative study of the origin of carbonate-hosted gem corundum deposits in Canada. Doctoral thesis, University of British Columbia, Vancouver, B.C., Canada. 259 p. Eliopoulos, D.G. & Economou-Eliopoulos, M. (2000): Geochemical and mineralogical characteristics of Fe-Ni- and bauxitic-laterite deposits of Greece. Ore Geology Reviews 16, 41-58. Ellis, D.E. (1978): Stability and phase equilibria of chloride and carbonate bearing scapolites at 750 °C and 4000 bar. Geochimica et Cosmochimica Acta 42, 1271-1281. Emmett, J.L., Stone-Sundberg, J., Guan, Y., & Sun, Z. (2017): The role of silicon in the color of gem corundum. Gems & Gemology 53, 42-47. Faryad, S.W. (2002): Metamorphic conditions and fluid compositions of scapolite-bearing rocks from the lapis lazuli deposit at Sare Sang, Afghanistan. Journal of Petrology 43, 725-747. Fleet, M.E., Liu, X., Harmer, S.L., & Nesbitt, H.W. (2005): Chemical state of sulfur in natural and synthetic lazurite by S K-edge XANES and X-ray photoelectron spectroscopy. Canadian Mineralogist 43, 1589-1603. 229  Fritsch, E. & Rossman, G.R. (1988): An update on colour in gems. Part 2: Colours involving multiple atoms and colour centers. Gems & Gemology 24, 3-15. Fritsch, E. & Rossman, G.R. (1987): An update on colour in gems. Part 1: Introduction and colours caused by dispersed metal ions. Gems & Gemology 23, 126-139. Garnier, V., Ohnenstetter, D., & Giuliani, G. (2004): L’aspidolite fluorée: role des évaporites dans la genèse du rubis des marbres de Nangimali (Azad-Kashmir, Pakistan). Comptes Rendus Geoscience 336, 1245-1253. Garnier, V., Giuliani, G., Ohnenstetter, D., Fallick, A.E., Dubessy, J., Banks, D., Hoang, Q.V., Lhomme, T., Maluski, H., Pêcher, A., Bakhsh, K.A., Long, P.V., Trinh, P.T., & Schwarz, D. (2008): Marble-hosted ruby deposits from Central and Southeast Asia: Towards a new genetic model. Ore Geology Reviews 34, 169-191. Gendron, A., Silverberg, N., Sundby, B., & Lebel, J. (1986): Early diagenesis of cadmium and cobalt in sediments of the Laurentian Trough. Geochimica et Cosmochimica Acta 50, 741-747. Genis, R. (2016): Retail gemstone price trends (1975-2015). The Gemstone Forecaster 34. http://www.preciousgemstones.com/gfspring16.html ; Retrieved 20 December, 2016. Gertzbein, P. (2003): The coloured gem potential of Baffin Island and Nunavut. Canadian Gemmologist 25, 10-17. Giletti, B.J. (1974): Studies in diffusion I: argon in phlogopite mica. In: Geochemical Transport and Kinetics. A.W. Hoffman, B.J. Giletti, H.S. Yoder, & R.A. Yund, editors. Carnegie Institution of Washington, publication 634, 107-115. 230  Giuliani, G., Dubessy, J., Ohnenstetter, D., Banks, D., Branquet, Y., Feneyrol, J., Fallick, A.E., & Martelat, J.-E. (2018): The role of evaporites in the formation of gems during metamorphism of carbonate platforms: a review. Mineralium Deposita 53, 1-20. Giuliani, G., Fallick, A.E., Boyce, A.J., Pardieu, V., & Pham, V.L. (2017): Pink and red spinels in marble: Trace elements, oxygen isotopes, and sources. Canadian Mineralogist 55, 743-761. Giuliani, G., Ohnenstetter, D., Fallick, A.E., Groat, L., & Fagan, A.J. (2014): The geology and genesis of gem corundum deposits. In Geology of Gem Deposits (L.A. Groat, ed.). Mineralogical Association of Canada short course volume 44 (29-112). Giuliani, G., Fallick, A.E., Garnier, V., France-Lanord, C., Ohnenstetter, D., & Schwarz, D. (2005): Oxygen isotope composition as a tracer for the origins of rubies and sapphires. Geology 33, 249-252. Goldsmith, J.R. & Newton, R.C. (1977): Scapolite-plagioclase stability relations at high pressures and temperatures in the system NaAlSi2O8-CaAl2Si2O8-CaCO3-CaSO4. American Mineralogist 62, 1063-1081. Graf, D.L. (1962): Minor element distribution in sedimentary carbonate rocks. Geochimica et Cosmochimica Acta 26, 849-856. Gregory, D.D., Large, R.R., Halpin, J.A., Baturina, E.L., Lyons, T.W., Wu, S., Danyushevsky, L., Sack, P.J., Chappaz, A., Maslennikov, V.V., & Bull, S.W. (2015): Trace element content of sedimentary pyrite in black shales. Economic Geology 110, 1389-1410.   231  Grew, E.S., Yates, M.G., Swihart, G.H., Moore, P.B. & Marquez, N. (1991): The paragenesis of serendibite at Johnsburg, New York, USA: an example of boron enrichment in the granulite facies. In Progress in metamorphic and magmatic petrology; a memorial volume in honor of D.S. Korzhinskiy (L.L. Perchuk, ed.). Cambridge University Press, Cambridge, England (247-285). Grice, J.D., Gault, R.A. & Waller, R. (1982): Spinel from Glencoe Island, Northwest Territories, Canada. Rocks & Minerals 57, 155-157. Grützner, T., Klemme, S., & Rohrbach, A. (2015) The effect of fluorine on the stability of humite-type minerals in the upper mantle and transition zone. Presented at the 25th Goldschmidt conference, August 16-21, 2015. Hålenius, U., Skogby, H., & Andreozzi, G.B. (2002): Influence of cation distribution on the optical absorption spectra of Fe3+-bearing spinel s.s.-hercynite crystals: Evidence for electron transitions in VIFe2+-VIFe3+ clusters. Physics and Chemistry of Minerals 29, 319-330. Hames, W.E. & Bowring, S.A. (1994): An empirical evaluation of the argon diffusion geometry in muscovite. Earth and Planetary Science Letters 124, 161-167. Hansen, M. (2008): A comparison of sapphire bearing and non-sapphire bearing lenses from the Lake Harbour Group, south-eastern Baffin Island. B.Sc. thesis, Carleton University, Ottawa, Ontario, Canada. 61 p. Harris, J.R., McGregor, R., & Budkewitsch, P. (2010): Geological analysis of hyperspectral data over southwest Baffin Island: methods for producing spectral maps that relate to variations in surface lithologies. Canadian Journal of Remote Sensing 36, 412-435. 232  Harrison, T.M., Célérier, J., Aikman, A.B., Hermann, J. & Heizler, M.T. (2009): Diffusion of 40Ar in muscovite. Geochimica et Cosmochimica Acta 73, 1039-1051. Hassan, I., Peterson, R.C., & Grundy, H.D. (1985): The structure of lazurite, ideally Na6Ca2(Al6Si6O24)S2, a member of the sodalite group. Acta Crystallographica C41, 827-832. Heggie, D. & Lewis, T. (1984): Cobalt in pore waters of marine sediments. Nature 311, 453-455. Herd, C.D.K., Peterson, R.C., & Rossman, G.R. (2000): Violet-coloured diopside from southern Baffin Island, Nunavut, Canada. Canadian Mineralogist 38, 1193-1199. Hogarth, D.D. (1979): Afghanite: New occurrences and chemical composition. Canadian Mineralogist 17, 47-52. Hogarth, D.D. (1971): Lapis lazuli near Lake Harbour, Southern Baffin Island, Canada. Canadian Journal of Earth Sciences 8, 1210-1217. Hogarth, D.D. & Griffin, W.L. (1980): Contact-metamorphic lapis lazuli: The Italian Mountain deposits, Colorado. Canadian Mineralogist 18, 59-70. Hogarth, D.D. & Griffin, W.L. (1978): Lapis lazuli from Baffin Island – a Precambrian meta-evaporite. Lithos 11, 37-60. Hogarth, D.D. & Griffin, W.L. (1975): Further data on lapis lazuli from Latium, Italy. Canadian Mineralogist 13, 89-90. Holland, T. & Powell, R. (1998): An internally consistent thermodynamic data set for phases of petrological interest. Journal of Metamorphic Geology 16, 309-343.  Holland, T. & Powell, R. (1991): A Compensated-Redlich-Kwong (CORK) equation for volumes and fugacities of CO2 and H2O in the range 1 bar to 50 kbar and 100-1600°C. Contributions to Mineralogy and Petrology 109, 265-273.  233  Housley, R. (2012): History of Cascade Canyon lapis and corundum deposits, San Gabriel Mountains, CA. Mindat.org article #1575, www.mindat.org/article.php/1575 - retrieved 26 February, 2017. Hughes, R.W. (1997): Ruby & sapphire. RWH Publishing, Boulder, Colorado, USA. 511 p. Hughes, R.W. & Pardieu, V. (2006): Moon over the Pamirs: Chasing ruby and spinel in Tajikistan. http://www.ruby-sapphire.com/tajikistan_ruby_and_spinel.htm retrieved March 10, 2017. Kievlenko, E.Y. (2003): Geology of gems, english edition. A. Soregaroli, editor. Ocean Pictures Ltd., Littleton, CO, USA. 468 pp. Kleišmantas, A. & Daukšytė, A. (2016): The influence of Vietnam and Sri Lanka spinel mineral chemical elements on colour. Chemija 27, 45-51. LeCheminant, A.N., Groat, L.A., Mortensen, J.K., Gertzbein, P., & Rohtert, W. (2005): Sapphires from Kimmirut, Baffin Island, Nunavut, Canada. Geochimica et Cosmochimica Acta 69, 10, supplement 280, Abstracts of the 15th annual V.M. Goldschmidt conference. Leeman, W.P. & Tonarini, S. (2001): Boron isotopic analysis of proposed borosilicate mineral reference samples. Geostandards Newsletter 25, 399-403. Lepage, L. & Davison, J.G. (2007): Exploration advances with ultraviolet LED technology; Beluga sapphire, Nunavut. Abstracts with program, Geological Association of Canada – Mineralogical Association of Canada joint annual meeting, Yellowknife, Northwest Territories, 2007. Lessing, P. & Grout, C.M. (1971): Haüynite from Edwards, New York. American Mineralogist 56, 1096-1100.  234  Locock, A.J. (2014): An Excel spreadsheet to classify chemical analyses of amphiboles following the IMA 2012 recommendations. Computers & Geosciences 62, 1-11. López, J.M.G., Bauluz, B., Fernández-Nieto, C., & Oliete, A.Y. (2005): Factors controlling the trace-element distribution in fine-grained rocks: the Albanian kaolinite-rich deposits of the Oliete Basin (NE Spain). Chemical Geology 214, 1-19. Ludwig, K.R. (2003): Isoplot 3.09: A geochronological toolkit for Microsoft Excel. Berkeley Geochronology Center (bgc.org), Special Publication no. 4. Ludwig, T., Marschall, H.R., Pogge von Strandmann, P.A.E., Shabaga, B.M., Fayek, M. & Hawthorne, F.C. (2011): A secondary ion mass spectrometry (SIMS) re-evaluation of B and Li isotopic compositions of Cu-bearing elbaite from three global localities. Mineralogical Magazine 75, 2485-2494. MacGregor, J.R., Grew, E.S., De Hoog, J.C.M., Harley, S.L., Kowalski, P.M., Yates, M.G. & Carson, C.J. (2013): Boron isotopic composition of tourmaline, prismatine, and grandidierite from granulite facies paragneisses in the Larsemann Hills, Prydz Bay, East Antarctica: Evidence for a non-marine evaporite source. Geochimica et Cosmochimica Acta, http://dx.doi.org/10.1016/j.gca.2013.05.030 Merlet, C. (1994): An accurate computer correction program for quantitative electron probe microanalysis. Microchimica Acta 114, 363-376. Moine, B., Sauvan, P., & Jarousse, J. (1981): Geochemistry of evaporite-bearing series: a tentative guide for the identification of metaevaporites. Contributions to Mineralogy and Petrology 76, 401-412. 235  Möller, A., ÓBrien, P.J., Kennedy, A. & Kröner, A. (2003): The use and abuse of Th-U ratios in the interpretation of zircon. In EGS-AGU-EUG joint assembly, Nice, France, 6-11 April 2003. Abstracts with program, abstract # 12113. Moore, T.P. & Woodside, R.W.M. (2014) The Sar-e-Sang lapis mines. Mineralogical Record 45, 281-336. Mora, C.I. & Valley, J.W. (1989): Halogen-rich scapolite and biotite: Implications for metamorphic fluid-rock interaction. American Mineralogist 74, 721-737. Mortensen, J.K., Ghosh, D., & Ferri, F. (1995): U-Pb age constraints of intrusive rocks associated with copper-gold porphyry deposits in the Canadian Cordillera. In Porphyry deposits of the northwestern Cordillera of North America. T.G. Schroeter, editor. Canadian Institute of Mining and Metallurgy, Special Volume 46, 142-158. Nockolds, S.R. (1954): Average chemical compositions of some igneous rocks. Bulletin of the Geological Society of America 65, 1007-1032. Palin, R.M., Weller, O.M., Waters, D.J., & Dyck, B. (2016): Quantifying geological uncertainty in metamorphic phase equilibria modelling; a Monte Carlo assessment and implications for tectonic interpretations. Geoscience Frontiers 7, 591-607. Palmer, M.R. (1991): Boron isotope systematics of hydrothermal fluids and tourmalines: A synthesis. Chemical Geology (Isotope Geoscience Section) 94, 111-121. Pardieu, V. (2014): Hunting for “jedi” spinels in Mogok. Gems & Gemology 50, 46-57. Pardieu, V. (2008): Tanzania, October 2007: a gemological safari, part 1. FieldGemology.org, http://www.fieldgemology.org/gemology%20tanzania%20ruby%20sapphire%20spinel%20tsavorite%20alexandrite%20emerald%20tunduru%20songea%20morogoro%20merelani.php ; retrieved June 10, 2015. 236  Pardieu, V., Hughes, R., & Boehm, E. (2008): Spinel: resurrection of a classic. InColor Magazine (Summer 2008 issue), 10-18. Parker, R.L. (1967): Composition of the Earth’s Crust. In: Michael Fleischer, editor. Data of Geochemistry, 6th ed. United States Government Printing Office, Washington. pp D1-D19. Pham, V.L., Pardieu, V., & Giuliani, G. (2013): Update on gemstone mining in Luc Yen, Vietnam. Gems & Gemology 49, 233-245. Plimer, I.R. (1977): The origin of the albite-rich rocks enclosing the cobaltian pyrite deposit at Thackaringa, N.S.W., Australia. Mineralium Deposita 12, 175-187. Pouchou, J.L. & Pichoir, F. (1985): PAP φ(ρZ) procedure for improved quantitative microanalysis. Microbeam Analysis Proceedings (ed., J.T. Armstrong), San Fransisco Press, California, USA, p. 104-106. Price, B.J. (1972): Minor elements in pyrites from the smithers map area, B.C. and exploration applications of minor element studies. Master’s Thesis, University of British Columbia, Vancouver, Canada, 270 pp. Renne, P.R., Swisher, C.C., Deino, A.L., Karner, D.B., Owens, T.L. & DePaolo, D.J. (1998): Intercalibration of standards, absolute ages and uncertainties in Ar-40/Ar-39 dating. Chemical Geology 145, 117-152. Rice, J.M. (1981) Phase equilibria involving humite minerals in impure dolomitic limestones: Part II – calculated stability of chondrodite and norbergite. Contributions to Mineralogy and Petrology 75, 205-223. Rivers, T. (2015): Tectonic setting and evolution of the Grenville Orogen: An assessment of progress over the last 40 years. Geoscience Canada 42, 77-124. 237  Robbins, G.A. (1972): Radiogenic argon diffusion in muscovite under hydrothermal conditions. Master’s thesis, Brown University. Roddick, J.C. (1987): Generalized numerical error analysis with application to geochronology and thermodynamics. Geochimica et Cosmochimica Acta 51, 2129-2135. Rubatto, D., Williams, I.S. & Günther, D. (1999): Trace-element characterization of metamorphic zircons. In 9th Annual V.M. Goldschmidt conference, L.P.I. Contribution 971, Lunar and Planetary Institute, Houston. Abstract 7111. Schropp, S.J. & Windom, H.L., eds. (1988): A guide to the interpretation of metal concentrations in estuarine sediments. Coastal Zone of Management Section, Florida Department of Environmental Regulation, Tallahassee, Florida, USA. 44 pp. Scott, D.J. (1997): Geology, U-Pb, and Pb-Pb geochronology of the Lake Harbour area, southern Baffin Island: implications for the Paleoproterozoic tectonic evolution of northeastern Laurentia. Canadian Journal of Earth Sciences 34, 140-155. Scott, D.J., Stern, R.A., St-Onge, M.R. & McMullen, S.M. (2002): U-Pb geochronology of detrital zircons in metasedimentary rocks from southern Baffin Island: implications for the Paleoproterozoic tectonic evolution of Northeastern Laurentia. Canadian Journal of Earth Sciences 39, 611-623. Scott, D.J., St-Onge, M.R., Wodicka, N., & Hanmer, S. (1997): Geology of Markham Bay-Crooks Inlet area, southern Baffin Island, Northwest Territories. In: Current Research 1997-C, Geological Survey of Canada, 157-166. Senoble, J.B. (2010): Beauty and rarity – a quest for Vietnamese blue spinels. InColor (Summer 2010 issue), 18-23. 238  Shaw, D.M. (1954): Trace elements in pelitic rocks: pt. I, Variation during metamorphism; pt. II, Geochemical relations. Geological Society of America Bulletin 65, 1151-1182. Shaw, T.J., Gieskes, J.M., & Jahnke, R.A. (1990): Early diagenesis in differing depositional environments: the response of transition metals in pore water. Geochimica et Cosmochimica Acta 54, 1233-1246. Shigley, J.E. & Stockton, C.M. (1984): ‘Cobalt-blue’ gem spinels. Gems & Gemology 20, 34-41. Shor, R. & Weldon, R. (2009): Ruby and sapphire production and distribution: a quarter century of change. Gems & Gemology 45, 236-259. Silva, K.K.M.W. & Siriwardena, C.H.E.R. (1988): Geology and the origin of the corundum-bearing skarn at Bakamuna, Sri Lanka. Mineralium Deposita 23, 186-190. Simonet, C., Fritsch, E. & Lasnier, B. (2008): A classification of gem corundum deposits aimed towards gem exploration. Ore Geology Reviews 34, 127-133. Skipton, D.R., St-Onge, M.R., Schneider, D.A., & McFarlane, C.R.M. (2016): Tectonothermal evolution of the middle crust in the Trans-Hudson Orogen, Baffin Island, Canada: Evidence from petrology and monazite geochronology of sillimanite-bearing migmatites. Journal of Petrology 57, 1437-1462. St-Onge, M.R., Scott, D.J., & Wodicka, N. (1999a): Geology, McKellar Bay, Nunavut; Geological Survey of Canada, Map 1981A, scale 1:100 000. St-Onge, M.R., Scott, D.J., & Wodicka, N. (1999b): Geology, Crooks Inlet, Nunavut; Geological Survey of Canada, Map 1984A, scale 1:100 000. St-Onge, M.R., Wodicka, N., & Lucas, S.B. (2000): Granulite- and amphibolite-facies metamorphism in a convergent-plate-margin setting: synthesis of the Quebec-Baffin segment of the Trans-Hudson Orogen. Canadian Mineralogist 38, 379-398. 239  St-Onge, M.R., Scott, D.J. & Wodicka, N. (2001): Geology, Meta Incognita Peninsula, Baffin Island, Nunavut. Geological Survey of Canada, Map 2009A. St-Onge, M.R., Wodicka, N., & Ijewliw, O. (2007): Polymetamorphic evolution of the Trans-Hudson Orogen, Baffin Island, Canada: Integration of petrological, structural and geochronological data. Journal of Petrology 48, 271-302. St-Onge, M.R., Van Gool, J.A.M., Garde, A.A., & Scott, D.J. (2009): Correlation of Archaean and Palaeoproterozoic units between northeastern Canada and western Greenland: constraining the pre-collisional upper plate accretionary history of the Trans-Hudson orogeny. Geological Society of London Special Publication 318, 193-235. St-Onge, M.R., Rayner, N.M., Liikane, D., & Chadwick, T. (2015): Mafic, ultramafic and layered mafic-ultramafic sills, Meta Incognita Peninsula, southern Baffin Island. In: Summary of Activities 2014, Canada-Nunavut Geoscience Office, 11-16. Stacey, J.S. & Kramer, J.D. (1975): Approximation of terrestrial lead isotope evolution by a two stage model. Earth and Planetary Science Letters 26, 207-221. Steiger, R.H. & Jäger, E. (1977): Subcommission on geochronology: convention on the use of decay constants in geo- and cosmochronology. Earth and Planetary Science Letters 36, 359-362. Stockdale, A., Davison, W., Zhang, H., & Hamilton-Taylor, J. (2010): The association of cobalt with iron and manganese (oxyhydr)oxides in marine sediment. Aquatic Geochemistry 16, 575-585. Swihart, G.H. & Moore, P.B. (1989): A reconnaissance of the boron isotopic composition of tourmaline. Geochimica et Cosmochimica Acta 53, 911-916.  240   Takahashi, Y., Hayasaka, Y., Morita, K., Kashiwabara, T., Nakada, R., Marcus, M.A., Kato, K., Tanaka, K. & Shimizu, H. (2015): Transfer of rare earth elements (REE) from manganese oxides to phosphates during early diagenesis in pelagic sediments inferred from REE patterns, X-ray absorption spectroscopy, and chemical leaching method. Geochemical Journal 49, 653-674.  Tauson, V.L., Goettlicher, J., Sapozhnikov, A.N., Mangold, S., & Lustenberg, E.E. (2012): Sulphur speciation in lazurite-type minerals (Na,Ca)8[Al6Si6O24](SO4,S)2 and their annealing products: a comparative XPS and XAS study. European Journal of Mineralogy 24, 133-152. Taylor, S.R. & McLennan, S.M. (1985) The continental crust: its composition and evolution. Blackwell Scientific Publications, Oxford, 312 pp. Themelis, T. (2008): Gems & Mines of Mogok. A & T Publishers, 356 p. Thériault, R.J., St-Onge, M.R. & Scott, D.J. (2001): Nd isotopic and geochemical signature of the Paleoproterozoic Trans-Hudson Orogen, southern Baffin Island, Canada: implications for the evolution of eastern Laurentia. Precambrian Research 108, 113-138. True North Gems Inc. (2007): True North Gems recovers 139 carat sapphire from new Baffin Island discovery. Press release, Dec. 17 2007, http://www.marketwired.com/press-release/true-north-gems-recovers-139-carat-sapphire-from-new-baffin-island-discovery-tsx-venture-tgx-803810.htm retrieved December 20, 2016. Turekian, K.K. & Wedepohl, K.H. (1961): Distribution of the elements in some major units of the Earth’s crust. Geological Society of America Bulletin 72, 175-191. 241  Turner, D.J., Groat, L.A., Rivard, B., & Belley, P.M. (2017): Reflectance spectroscopy and hyperspectral imaging of sapphire-bearing marble from the Beluga occurrence, Baffin Island, Nunavut. Canadian Mineralogist 55, 787-797. Tuttle, M.L., Klett, T.R., Richardson, M. & Breit, G.N. (2000): Geochemistry of two interbeds in the Pennsylvanian Paradox Formation, Utah and Colorado – A record of deposition and diagenesis of repetitive cycles in a marine basin. US Geological Survey Bulletin 2000-N, 86 p. van Hinsberg, V.J., Henry, D.J., & Marschall, H.R. (2011): Tourmaline: an ideal indicator of its host environment. Canadian Mineralogist 49, 1-16. Van Velthuizen, J. (1993): The Parker mine, Notre Dame du Laus, Quebec. Mineralogical Record 24, 369-373. Villa, I.M. (2010): Disequilibrium textures versus equilibrium modelling: geochronology at the crossroads. In: Advances in interpretation of geological processes: refinement of multi-scale data and integration in numerical modelling. M.I. Spalla, A.M. Marotta, G. Gosso, editors. Geological Society of London Special Publication 332, 1-15. Wang, K.K., Graham, I.T., Lay, A., Harris, S.J., Cohen, D.R., Voudoris, P., Belousova, E., Giuliani, G., Fallick, A.E., & Greig, A. (2017): The origin of a new pargasite-schist hosted ruby deposit from Paranesti, Northern Greece. Canadian Mineralogist 55, 535-560. Warren, J.K. (2016): Meta-evaporites. In: Evaporites: a geological compendium. Springer, pp. 1375-1478. 242  Wenger, L.M. & Baker, D.R. (1986): Variations in organic geochemistry of anoxic-oxic black shale-carbonate sequences in the Pennsylvanian of the Midcontinent, U.S.A. Advances in Organic Geochemistry 10, 85-92. Wilson, B.S. (2014): Coloured gemstones from Canada. In: Lee Groat (editor), Geology of Gem Deposits. Mineralogical Association of Canada, Short Course 44, 375-405. Yakymchuk, C. & Szilas, K. (2018): Corundum formation by metasomatic reactions in Archean metapelite, SW Greenland: Exploration vectors for ruby deposits within high-grade greenstone belts. Geoscience Frontiers 9, 727-749. Yardley, B.W.D. & Graham, J.T. (2002): The origins of salinity in metamorphic fluids. Geofluids 2, 249-256. Appendix A  Appendix A1: Argon data collected from phlogopite sample “A” (Beluga occurrence).  Step 40Ar/39Ar 2σ 38Ar/39Ar 2σ 37Ar/39Ar 2σ 36Ar/39Ar 2σ Ca/K Cl/K %40Ar (atm) f 39Ar 40Ar*/39ArK 2σ Age 2σ Phlogopite-A                1 86.792 0.008 0.083 0.135 0.041 0.130 0.139 0.040 0.160 0.010 47.07 0.76 45.993 1.675 1564.28 38.11 2 55.846 0.007 0.032 0.070 0.008 0.231 0.022 0.077 0.032 0.003 11.64 2.34 49.408 0.601 1640.31 13.11 3 50.105 0.005 0.018 0.087 0.001 0.767 0.003 0.142 0.004 0.001 1.96 8.27 49.185 0.272 1635.45 5.96 4 55.968 0.008 0.052 0.084 0.005 1.152 0.023 0.103 0.020 0.008 11.88 0.96 49.380 0.797 1639.71 17.38 5 49.880 0.005 0.017 0.098 0.001 1.319 0.002 0.392 0.004 0.001 1.27 5.39 49.307 0.363 1638.12 7.93 6 49.965 0.005 0.016 0.053 0.002 0.405 0.003 0.144 0.007 0.000 1.73 6.91 49.160 0.288 1634.90 6.29 7 49.821 0.006 0.014 0.155 0.001 1.245 0.001 1.207 0.005 0.000 0.59 5.57 49.590 0.469 1644.27 10.20 8 49.877 0.004 0.017 0.057 0.002 0.412 0.003 0.118 0.007 0.001 1.98 6.13 48.950 0.252 1630.30 5.52 9 49.820 0.005 0.016 0.077 0.001 0.361 0.002 0.182 0.005 0.000 1.36 8.99 49.205 0.269 1635.89 5.89 10 49.519 0.005 0.016 0.061 0.002 0.432 0.002 0.271 0.007 0.000 1.16 8.21 49.005 0.294 1631.51 6.43 11 49.787 0.006 0.016 0.115 0.002 0.403 0.001 0.430 0.006 0.000 0.86 6.22 49.419 0.347 1640.56 7.57 12 49.642 0.005 0.014 0.098 0.001 0.954 0.001 0.717 0.004 0.000 0.61 6.12 49.402 0.346 1640.20 7.54 13 49.536 0.005 0.015 0.084 0.002 0.386 0.001 0.769 0.007 0.000 0.47 6.86 49.364 0.323 1639.35 7.05 14 49.463 0.005 0.014 0.096 0.002 0.485 0.001 0.709 0.007 0.000 0.46 7.16 49.295 0.312 1637.84 6.81 15 49.466 0.006 0.015 0.124 0.000 2.291 0.001 0.892 0.002 0.000 0.47 4.67 49.293 0.377 1637.81 8.23 16 49.200 0.006 0.014 0.083 0.001 0.526 0.001 0.326 0.005 0.000 0.69 7.61 48.922 0.314 1629.69 6.88 17 49.049 0.004 0.015 0.085 0.001 0.787 0.001 0.618 0.003 0.000 0.36 7.83 48.931 0.249 1629.88 5.46 Total/Avg 50.227 0.001 0.017 0.011 0.004 0.029 0.003 0.024  0.001  100.00 49.205 0.047    J = 0.030003 ± 0.000060, Volume 39ArK = 442.9 × 10-13 cm3           244      Appendix A2: Argon data collected from phlogopite sample “A” (Beluga occurrence).  Step 40Ar/39Ar 2σ 38Ar/39Ar 2σ 37Ar/39Ar 2σ 36Ar/39Ar 2σ Ca/K Cl/K %40Ar (atm) f 39Ar 40Ar*/39ArK 2σ Age 2σ Phlogopite-B                1 60.154 0.009 0.037 0.312 0.007 0.580 0.040 0.133 0.227 0.004 19.73 0.48 48.175 1.634 1613.22 36.16 2 55.607 0.005 0.018 0.126 0.001 1.354 0.007 0.117 0.033 0.001 3.93 3.43 53.300 0.389 1723.25 8.10 3 52.427 0.005 0.016 0.072 0.001 0.396 0.002 0.128 0.043 0.000 1.32 6.74 51.619 0.296 1687.90 6.29 4 51.976 0.006 0.016 0.058 0.000 3.278 0.003 0.128 0.005 0.000 1.74 9.57 50.953 0.314 1673.70 6.72 5 50.200 0.004 0.015 0.053 0.000 1.188 0.001 0.166 0.006 0.000 0.71 16.03 49.730 0.226 1647.33 4.90 6 51.229 0.011 0.016 0.258 0.003 1.742 0.003 0.420 0.085 0.000 1.93 1.52 50.124 0.679 1655.87 14.67 7 49.842 0.005 0.014 0.061 0.000 1.596 0.001 0.308 0.009 0.000 0.48 15.88 49.488 0.257 1642.07 5.60 8 49.954 0.005 0.015 0.068 0.000 1.082 0.001 0.450 0.013 0.000 0.33 9.75 49.675 0.242 1646.13 5.25 9 50.105 0.007 0.015 0.091 0.001 0.922 0.001 0.735 0.026 0.000 0.56 7.68 49.709 0.399 1646.86 8.67 10 50.267 0.006 0.013 0.057 0.000 2.067 0.001 0.555 0.004 0.000 0.32 12.82 49.989 0.339 1652.94 7.34 11 50.588 0.006 0.015 0.099 0.000 1.444 0.001 0.300 0.000 0.000 0.60 7.97 50.167 0.294 1656.79 6.34 12 50.839 0.007 0.014 0.104 0.000 1.859 0.001 0.512 0.000 0.000 0.41 6.28 50.512 0.349 1664.23 7.50 13 51.765 0.011 0.014 0.222 -0.001 2.903 0.004 0.586 0.000 0.000 2.05 1.84 50.586 0.850 1665.81 18.28 Total/Avg 50.703 0.001 0.015 0.011 0.007 0.012 0.002 0.036  0.000  100.00 49.617 0.053    J = 0.030003 ± 0.000060, Volume 39ArK = 269.57 × 10-13 cm3             245  Appendix A3: Argon data collected from muscovite sample “A” (Beluga occurrence).  Step 40Ar/39Ar 2σ 38Ar/39Ar 2σ 37Ar/39Ar 2σ 36Ar/39Ar 2σ Ca/K Cl/K %40Ar (atm) f 39Ar 40Ar*/39ArK 2σ Age 2σ Muscovite-A                1 33.725 0.006 0.094 0.038 0.027 0.077 0.050 0.038 0.110 0.016 43.09 3.95 19.197 0.567 820.57 19.46 2 27.641 0.005 0.063 0.075 0.216 0.018 0.020 0.056 0.875 0.010 21.03 6.02 21.845 0.350 909.28 11.44 3 39.520 0.004 0.069 0.043 0.467 0.018 0.008 0.069 1.891 0.012 5.84 14.76 37.277 0.240 1354.19 6.13 4 42.220 0.005 0.019 0.050 0.004 0.113 0.001 0.395 0.017 0.001 0.50 14.25 42.057 0.249 1472.36 5.97 5 39.617 0.004 0.018 0.055 0.002 0.267 0.001 0.561 0.009 0.001 0.47 17.03 39.474 0.207 1409.47 5.14 6 43.683 0.006 0.013 0.098 0.002 0.489 0.000 0.741 0.006 0.000 0.32 15.75 43.593 0.285 1508.75 6.68 7 43.588 0.005 0.013 0.118 0.001 0.785 0.000 1.761 0.004 0.000 0.19 12.88 43.553 0.275 1507.81 6.45 8 44.105 0.005 0.011 0.201 0.001 2.321 0.000 2.832 0.002 -0.001 0.20 7.36 44.070 0.347 1519.90 8.09 9 44.153 0.007 0.012 0.122 0.003 0.419 0.001 0.999 0.010 0.000 0.64 6.02 43.920 0.414 1516.40 9.68 10 44.791 0.010 0.016 0.252 0.007 0.709 0.006 0.657 0.030 0.000 3.99 1.98 43.052 1.268 1496.02 29.94 Total/Avg 40.954 0.001 0.029 0.012 0.187 0.004 0.005 0.021  0.004  100.00 43.682 0.052    J = 0.030000 ± 0.000062, Volume 39ArK = 311.75 × 10-13 cm3               246  Appendix A4: Argon data collected from muscovite sample “B” (Beluga occurrence).  Step 40Ar/39Ar 2σ 38Ar/39Ar 2σ 37Ar/39Ar 2σ 36Ar/39Ar 2σ Ca/K Cl/K %40Ar (atm) f 39Ar 40Ar*/39ArK 2σ Age 2σ Muscovite-B                1 10.793 0.009 0.060 0.026 0.009 0.082 0.023 0.040 0.286 0.010 63.18 13.23 3.937 0.280 201.43 13.55 2 55.429 0.006 0.056 0.025 0.113 0.037 0.047 0.022 3.691 0.008 24.93 4.88 41.568 0.393 1460.62 9.46 3 45.162 0.006 0.021 0.053 0.014 0.094 0.004 0.058 0.468 0.002 2.81 9.76 43.794 0.295 1513.46 6.91 4 43.847 0.008 0.013 0.046 0.000 1.104 0.001 0.411 0.010 0.000 0.53 18.74 43.511 0.344 1506.81 8.08 5 43.892 0.010 0.013 0.080 0.001 0.664 0.001 0.260 0.030 0.000 0.62 13.38 43.516 0.437 1506.94 10.26 6 43.660 0.008 0.013 0.092 0.000 2.164 0.001 0.559 0.011 0.000 0.41 15.62 43.376 0.356 1503.64 8.36 7 44.058 0.005 0.013 0.069 0.001 0.419 0.000 0.578 0.019 0.000 0.25 20.52 43.841 0.252 1514.55 5.88 8 46.009 0.014 0.014 0.369 0.004 0.847 0.003 0.978 0.136 0.000 1.79 1.76 45.082 1.028 1543.32 23.65 9 46.587 0.009 0.013 0.355 0.002 1.304 0.005 0.349 0.067 -0.001 3.37 2.10 44.911 0.677 1539.39 15.60 Total/Avg 40.213 0.001 0.022 0.009 0.152 0.002 0.006 0.014  0.002  100.00 43.607 0.064    J = 0.030000 ± 0.000062, Volume 39ArK = 198.62 × 10-13 cm3            

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0378174/manifest

Comment

Related Items