Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The Renard 65 kimberlites : emplacement-related processes in Kimberley-type pyroclastic kimberlites Gaudet, Matthew A. 2016

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

Item Metadata

Download

Media
24-ubc_2017_february_gaudet_matthew.pdf [ 13.41MB ]
Metadata
JSON: 24-1.0340671.json
JSON-LD: 24-1.0340671-ld.json
RDF/XML (Pretty): 24-1.0340671-rdf.xml
RDF/JSON: 24-1.0340671-rdf.json
Turtle: 24-1.0340671-turtle.txt
N-Triples: 24-1.0340671-rdf-ntriples.txt
Original Record: 24-1.0340671-source.json
Full Text
24-1.0340671-fulltext.txt
Citation
24-1.0340671.ris

Full Text

  The Renard 65 kimberlites: emplacement-related processes in Kimberley-type pyroclastic kimberlites  by  Matthew A. Gaudet B.Sc. Hons, Dalhousie University, 2013   A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in The Faculty of Graduate and Postdoctoral Studies (Geological Sciences)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  December 2016   ©Matthew A. Gaudet, 2016 i  Abstract The Renard 65 pipe is located in the Otish Mountains, Quebec, Canada. It is one of nine diamondiferous kimberlite pipes in the ~ 640 Ma Renard cluster and is the largest of four pipes in the Renard Mine reserve. Detailed characterizations of the petrographic and compositional features of these pipe-infilling kimberlite rock types supports their classification into three geological units: Kimb65a, Kimb65b, and Kimb65d. These pipe-infilling kimberlites are interpreted to represent the solidified products of two separate magmatic events: Phase A containing Kimb65a, and Phase B containing Kimb65b and Kimb65d. This research demonstrates that the interclast matrix modal mineralogy (diopside + phlogopite + serpentine) in pyroclastic rock types in the Renard 65 kimberlites are inconsistent with origins by hydrothermal alteration involving hydrous meteoric fluids. Detailed investigation of the reactions between granitic and gneissic crustal xenolith lithologies and their host kimberlites, suggests that reactions occur at both magmatic and subsolidus temperatures involving significant volumetric proportions of xenoliths. The assimilation of crustal xenoliths, and contamination of the kimberlite magmas primarily by Si, are demonstrated to result in enhanced degassing of magmatic volatiles during emplacement and stabilization of the hybrid groundmass assemblage diopside + phlogopite + serpentine over the non hybrid groundmass assemblage calcite + phlogopite + serpentine. It is thus interpreted that the spatial distribution of transitional to Kimberley-type pyroclastic kimberlite rock types, which are characterized by diopside-rich and calcite-poor matrix assemblages as observed in the Renard 65 pipe and other similar pipes, is a function of crustal xenolith distribution in the magma during emplacement. This model not only accounts for the features of Kimberley-type pyroclastic kimberlite rock types, but also the spatial distribution of these rock types in numerous pipes which is often not consistent with lateral textural gradations as has been previously proposed. These results further indicate that the different mineralogy and textures of Fort-à-la-Corne-type pyroclastic kimberlites with respect to Kimberley-type pyroclastic kimberlites may be a consequence of not only the structural controls imparted by the host rock lithology with implications for emplacement-related processes, but also the absence of contamination of the magma by silicic crustal xenoliths.     ii  Preface This dissertation is original, unpublished, independent work by the author, M. Gaudet.                          iii  Table of Contents  Abstract .......................................................................................................................................................... i Preface .......................................................................................................................................................... ii Table of Contents ......................................................................................................................................... iii List of Tables ................................................................................................................................................. v List of Figures ............................................................................................................................................... vi List of Supplementary Materials ................................................................................................................ viii Glossary ........................................................................................................................................................ ix Acknowledgements .....................................................................................................................................xiii 1. Introduction .......................................................................................................................................... 1 1.1 Statement of the Research Problem ............................................................................................. 1 1.2 Kimberlite Defined ........................................................................................................................ 3 1.3 Kimberlite Classified ...................................................................................................................... 5 1.4 Kimberlite Emplacement ............................................................................................................ 13 1.5 Regional and Local Geology of the Renard Kimberlites .............................................................. 15 1.6 Research Objectives .................................................................................................................... 19 2. Methods .............................................................................................................................................. 20 2.1 Core Logging and Polished Slab Descriptions ............................................................................. 20 2.2 Kimberlite Petrography and Scanning Electron Microscopy ...................................................... 21 2.3 Mineral Chemistry ....................................................................................................................... 22 3. Results ................................................................................................................................................. 23 3.1 Geological Unit Summaries ......................................................................................................... 23 3.1.1 Kimb65a .............................................................................................................................. 25 3.1.2 Kimb65b .............................................................................................................................. 33 3.1.3 Kimb65c............................................................................................................................... 41 3.1.4 Kimb65d .............................................................................................................................. 46 3.1.5 CR, CCR, CRB ........................................................................................................................ 52 3.2 Mineral Chemistry ....................................................................................................................... 54 3.2.1 Olivine ................................................................................................................................. 54 3.2.2 Carbonate ............................................................................................................................ 55 3.2.3 Phlogopite ........................................................................................................................... 57  iv  3.2.4 Spinel ................................................................................................................................... 59 3.2.5 Perovskite ............................................................................................................................ 65 3.2.6 Apatite ................................................................................................................................. 67 3.2.7 Pyroxenes and Pectolite ...................................................................................................... 68 3.2.8 Amphibole ........................................................................................................................... 70 4. Renard 65 Rock Classification ............................................................................................................. 71 4.1 Comparison of Features with Group I Kimberlites ..................................................................... 71 4.2 Mineralogical Classification of Renard 65 Kimberlites ............................................................... 74 4.3 Textural-Genetic Classification ................................................................................................... 76 5. Discussion ............................................................................................................................................ 78 5.1 The Origin and Ascent of Kimberlite Magmas ............................................................................ 78 5.2 Kimberlite Emplacement: Processes of Diatreme Formation and Infill ...................................... 79 5.3 Current Models on the Origin of the Interclast Matrix in KPKs .................................................. 84 5.4 Macroscopic Features of Crustal Xenoliths in KPKs .................................................................... 90 5.5 Reactions Involving Crustal Xenoliths and Host Kimberlite in KPKs ........................................... 91 5.6 Crustal Xenoliths in Renard 65 .................................................................................................... 93 5.6.1 Reactions between Crustal Xenoliths in KPK rock types in Kimb65a .................................. 94 5.6.2 Reactions between Crustal Xenoliths in HK, HKt, and KPKt Rock Types in Kimb65b and Kimb65d .............................................................................................................................. 97 5.7 Degree of Assimilation of Crustal Xenoliths in Kimberlites ...................................................... 107 5.8 Effects of Crustal Xenolith Assimilation on Volatile Fluid Exsolution in Kimberlite Melt ......... 109 5.9 Perovskite Alteration by CO2-rich Fluids ................................................................................... 114 5.10 Oxygen Fugacity ........................................................................................................................ 119 5.11 Geological Modelling of the Renard 65 Pipe ............................................................................ 125 5.11.1 Identification of Kimberlite Phases for Geological Modelling .......................................... 125 5.11.2 Phases of Kimberlite in Renard 65 .................................................................................... 126 5.11.3 The Emplacement of Renard 65 Kimberlites and the Origin of the Interclast Matrix ...... 133 5.11.4 Renard 65 Geological Model ............................................................................................. 136 5.12 Contrasting Texture and Mineralogy of KPK and FPK Pipes ..................................................... 136 6. Conclusions ....................................................................................................................................... 139 References ................................................................................................................................................ 142   v  List of Tables Table 3.1 - Renard 65 Geological Unit Summary ........................................................................................ 24 Table 3.2 – Compositions of olivine macrocrysts in coherent rocks from Kimb65b .................................. 55 Table 3.3 – Averaged calcite compositions ................................................................................................. 56 Table 3.4 – Averaged phlogopite compositions .......................................................................................... 58 Table 3.5 – Averaged spinel compositions ................................................................................................. 60 Table 3.6 – Averaged perovskite compositions .......................................................................................... 66 Table 3.7 – Averaged apatite compositions ................................................................................................ 67 Table 3.8 – Averaged compositions of diopside, aegirine and pectolite .................................................... 68 Table 3.9 – Compositions of amphibole in crustal xenoliths in volcaniclastic rock types in Kimb65a ....... 70 Table 4.1 Mineralogical Classification of Renard 65 Geological Units ........................................................ 75 Table 5.1 – Modelled hybrid kimberlite melt compositions for minimum and maximum initial SiO2 kimberlite melt compositions resulting from various degrees of assimilation of different modal abundances of crustal xenoliths. .............................................................................................................. 111 Table 5.2 – Calculated minimum and maximum increases in wt% SiO2 for the Renard 65 geological units .................................................................................................................................................................. 112              vi  List of Figures Figure 1.1 – Textures of Fort-à-la-Corne type pyroclastic kimberlites and Kimberley-type pyroclastic kimberlites .................................................................................................................................................... 2 Figure 1.2 – Conceptual framework for the description of kimberlite components .................................... 7 Figure 1.3 – Size and abundance descriptors for compound clasts and crystals .......................................... 8 Figure 1.4 – 5 step systematic framework for the description, classification, and subsequent    interpretation of kimberlites ...................................................................................................................... 10 Figure 1.5 – Classic model of an idealized kimberlite magmatic system .................................................... 13 Figure 1.6 – Geology of the Superior Province ........................................................................................... 15 Figure 1.7 – Geological setting of the Renard cluster ................................................................................. 16 Figure 1.8 – Geological Model of the Renard 65 Pipe ................................................................................ 18 Figure 3.1 – Kimb65a: macroscopic textures, magmaclasts and xenoliths ................................................ 26 Figure 3.2 – Kimb65a: microscopic features of magmaclasts, olivine and interclast matrix ...................... 28 Figure 3.3 – Kimb65a: groundmass mineral assemblages in well preserved magmaclasts ....................... 30 Figure 3.4 – Kimb65a: groundmass mineral assemblages in poorly preserved magmaclasts.................... 30 Figure 3.5 – Kimb65a: groundmass mineral textures ................................................................................. 31 Figure 3.6 – Kimb65b: macroscopic kimberlite rock texture ...................................................................... 34 Figure 3.7 – Kimb65b: xenolith reaction mineral assemblages .................................................................. 35 Figure 3.8 – Kimb65b: coherent groundmass mineral assemblage ............................................................ 37 Figure 3.9 – Kimb65b: transitional coherent groundmass mineral assemblage ........................................ 37 Figure 3.10 – Kimb65b: groundmass and interclast matrix mineral textures ............................................ 39 Figure 3.11 – Kimb65c: megascopic and macroscopic features ................................................................. 42 Figure 3.12 – Kimb65c: megascopic cross-cutting relationships ................................................................ 43 Figure 3.13 – Kimb65c: groundmass mineral assemblage .......................................................................... 45 Figure 3.14 – Kimb65c: phenocryst and groundmass mineral textures ..................................................... 45 Figure 3.15 – Kimb65d: kimberlite texture and xenolith reaction assemblages ........................................ 47 Figure 3.16 – Kimb65d: groundmass mineral assemblage ......................................................................... 49 Figure 3.17 – Kimb65d: groundmass and interclast matrix mineral textures. ........................................... 51 Figure 3.18 – CR, CCR and CRB rock textures ............................................................................................. 53 Figure 3.19 – Graphical illustrations of spinel compositions ...................................................................... 61 Figure 3.20 – Spinel core compositions general ......................................................................................... 63 Figure 3.21 – Spinel core compositions detailed ........................................................................................ 64  vii  Figure 4.1 – Al2O3 vs. TiO2 and Al2O3 vs. FeO diagrams for micas in Renard 65 ......................................... 73 Figure 5.1 – KPK and KPKt rock types in Kimb65a ...................................................................................... 94 Figure 5.2 – Occurrence of eckermannite in crustal xenoliths in Kimb65a ................................................ 95 Figure 5.3 – Chloritization of biotite in crustal xenoliths in Kimb65a ......................................................... 96 Figure 5.4 – Macroscopic Texture of HKt rock types in Kimb65b ............................................................... 98 Figure 5.5 – Illustration of in-situ magmatic reactions in Kimb65b and Kimb65d ..................................... 99 Figure 5.6 – Illustration of subsolidus bimetasomatism in Kimb65b and Kimb65d ................................. 102 Figure 5.7 – Chemographic Na, K, Ca – Mg, Fe – Si diagram for Kimb65b and Kimb65d ......................... 103 Figure 5.8 – Aegirine-diopside zoning in Kimb65b and Kimb65d ............................................................. 104 Figure 5.9 – H2O-CO2 dependency ................................................................. Error! Bookmark not defined. Figure 5.10 – Solubility of CO2 as a function of both pressure and Si content in the melt ...................... 113 Figure 5.11 – Perovskite equilibria with ilmenite-group minerals ............................................................ 117 Figure 5.12 – Common redox buffering assemblages .............................................................................. 119 Figure 5.13 – Oxygen fugacity ƒO2 of the Renard 65 kimberlites .............................................................. 122 Figure 5.14 – Morphological features of perovskite in Kimb65b and Kimb65d ....................................... 124 Figure 5.15 – Cross-cutting relationships between Phase A and Phase B ................................................ 128 Figure 5.16 – Morphological differences between groundmass spinel, perovskite and phlogopite in Kimb65b and Kimb65d .............................................................................................................................. 130 Figure 5.17 – Correlations between crustal xenolith modal abundances and kimberlite rock textures for kimberlites in multiple pipes from Gahcho Kue (Tuzo, Hearne, 5034), Renard 3 & Renard 65) .............. 135 Figure 5.18 – Revised Geological Model for Renard 65 ............................................................................ 136 Figure 5.19 – Schematic representation of the geology of FPK, RVK and KPK pipes ................................ 137         viii  List of Supplementary Materials Appendix A – Renard 65 drill core logs Appendix B – Renard 65 polished slab descriptions Appendix C – Renard 65 complete mineral chemistry                      ix  Glossary Kimberlite terminology definitions are quoted directly from “A Glossary of Kimberlite and Related Terms” - Scott Smith et al. (in press). Autoclast: A fragment of solidified magma formed by in-situ, non-explosive fragmentation during the movement of partly cooled magma (lava) and occurring within the same magma. Autoclasts typically form during effusive eruptions. Not synonymous with autolith. Autolith: An inclusion of earlier, previously consolidated kimberlite. Autoliths are typically angular or irregular in shape with broken components (e.g. olivine crystal, xenolith) exposed at the outer surface, showing that they formed by the brittle fragmentation of earlier consolidated kimberlite. Autoliths can become rounded and may resemble, and be difficult to differentiate from, magmaclasts. Compound Clasts: A general term for clasts composed of two or more constituents, typically an assemblage of crystals or grains of unspecified character and origin, which is distinct from surrounding material. Compound clasts include kimberlitic, mantle and crustal types. Epiclast: Particles or detritus created from pre-existing rocks by surface processes such as chemical and/or physical weathering, abrasion, erosion and mass wasting. Geological Unit: A part of a whole. A part or subdivision of rock which has unifying characteristics that are distinct from adjacent parts. A geological unit is distinctive enough from adjacent parts to be meaningfully separated or delineated for description and interpretation. Kimberlite units could be defined using features such as lithological characteristics or other criteria such as geophysics or changes in diamond grade. A kimberlite unit may with further interpretation be understood to be a phase of kimberlite. Multiple kimberlite units may together form a phase of kimberlite. Groundmass: In igneous rocks the term groundmass described the fine-grained minerals which crystallize relatively rapidly or glass, other amorphous or cryptocrystalline material that quenches from the late-stage melt between any pre-existing phenocrysts, typically during or after final emplacement. Interclast Matrix: The material occurring between clasts (sensuo lato) or the framework grains (see interstitial matrix).  x  Interstitial Matrix: A lithologic or petrographic term denoting the interstitial material occurring between or “supporting” larger crystals, fragments or particles. It is the background material in which larger particles and fragments occur. Interstitial matrix may comprise any material: clastic, crystalline, melt, volatile or other solidification products. Kimberlitic Compound Clast: A type of compound clast comprised entirely or partly of kimberlitic constituents. Kimberlitic compound clasts include magmaclasts, lithic clasts, and accretionary clasts. Lithic Clast: A clast (or fragment) formed by brittle fragmentation of lithified rocks. In volcanology, this includes both volcanic and non-volcanic rock types. Lithic clasts in volcanic rocks are usually angular, but may be rounded during eruption. The two main processes that produce lithic fragments in volcanic rocks are (a) volcanic eruptions and (b) surface weather and erosion. The latter are commonly termed epiclasts. In kimberlites, lithic clasts can also result from non-volcanic intrusive processes. Lithic clasts include autoclasts, autoliths, and epiclasts. Macrocryst: A non-genetic term used to describe typically large, anhedral and commonly round crystals which occur in kimberlites and related rocks. Macrocrysts in kimberlites typically have a wide size range (>0.5 mm) and are commonly medium- to coarse-grained (2-8 mm) or very coarse-grained (8-16 mm) and can therefore be seen with the unaided eye. The macrocrysts are set in a finer-grained matrix which gives most kimberlites their characteristic macroscopic inequigranular or macrocrystic texture. Olivine macrocrysts are a dominant and characteristic component of typical coherent kimberlite, generally forming ~25 volume %. In addition to olivine, many kimberlites contain other less common but distinctive macrocrysts; these include pyrope garnet, Mg-ilmenite, Cr-spinel, chrome diopside, and other types of clinopyroxene, orthopyroxene, phlogopite, zircon, and diamond. In many instances, the olivine and associated macrocrysts in kimberlite can be shown to be mantle-derived xenocrysts, and by association in other instances they are inferred to be mantle-derived.     xi  Magmaclasts: A general non-genetic descriptive interim term for a solidified fragment of fluidal magma produced by any process. The term can be replaced by more specific terms when further work permits more advanced interpretation (e.g. melt segregation, melt-bearing pyroclasts).  Magmaclast described a physically distinct or definable body of solidified magma comprising former melt plus any entrained solids (xenoliths, xenocrysts, macrocrysts, phenocrysts) as well as dissolved and exsolved volatile-rich fluids. The fluidal shapes of magmaclasts indicate derivation from magmas with low viscosity melts, and that they were hot immediately prior to formation and shaped by surface tension processes prior to solidification. Melt-Bearing Pyroclast: a kimberlite pyroclast containing solidified kimberlite magma, produced during kimberlite volcanic eruptions. Melt-Segregation: One variety of segregation comprising discrete bodies of melt occurring either within a fluid or melt host. The particular types of melt segregations relevant to kimberlites include (i) segregation of melt in fluid, and (ii) segregations of melt in melt.  Mesostasis: The part of the groundmass resulting from the crystallization or consolidation (as glass, other amorphous or crystalline material) of the final fraction of melt between existing crystals including early groundmass. The last-formed interstitial minerals or other material of an igneous rock. Phenocryst: Well-formed, typically euhedral-to-subhedral crystals inferred to be early liquidus phases which crystallized from the melt. Phenocrysts form earlier and at higher pressures and temperatures, i.e. at greater depths, than groundmass minerals. Phenocrysts are typically larger crystals set in a finer-grained or glassy groundmass or in glass and are presumed to have grown from the liquid now represented by the groundmass or glass. Pyrocryst: A melt-free crystal completely separated from the host melt by primary pyroclastic processes. Includes crystals occurring within the host magmatic fluids but separated from the melt. Pyrocrysts can be common in certain pyroclastic kimberlites because the high proportion of crystals in the volatile-rich, low-viscosity magmas are readily separated from the melt during pyroclastic processes. Pyrocrysts can occur subsurface or be liberated at the Earth’s surface.    xii  Xenocryst: A foreign accidental or extraneous crystal occurring in an igneous or volcanic rock. Xenocrysts derive from pre-existing genetically unrelated rocks (crustal or mantle) or surface materials. May also be derived from the disaggregation of a xenolith. Some may resemble phenocrysts. Xenolith: A fragment of country rock within an igneous or volcanic rock. A foreign, accidental or extraneous rock inclusion. A fragment of pre-existing genetically unrelated country rock derived from the wall rock during ascent or emplacement.                    xiii  Acknowledgements  Thank-you to my supervisor Maya Kopylova for providing me with an opportunity to join the Diamond Exploration Lab at the University of British Columbia to pursue my interests in economic geology and petrography by facilitating this research program with the Canadian diamond mining company Stornoway Diamond Corporation. Before arriving at UBC, I had little knowledge of kimberlites or diamond exploration/mining, but with Maya’s leadership and guidance I have overcome every hurdle to arrive at the pinnacle of kimberlite geology. Thank you for pushing me to be the very best that I can be.  I would like to thank the entire geological staff at the Stornoway Diamond Corporation exploration office in North Vancouver, for not only providing me with a challenging and thought provoking project, but also with unparalleled training and work experience on a kimberlite drill-core logging program. The experience working with Stornoway was instrumental to my ability to carry out this research, and I am immensely grateful for the opportunity. I will always appreciate the opportunities I’ve had to share thoughts and ideas and have engaging discussions with the Stornoway geological staff. Thank you for encouraging critical thought, skepticism, and instilling a sense of wonder in me for kimberlite geology.  I would also like to thank the many professionals in the diamond industry who provided constructive feedback and shared valuable insights from their experiences working on kimberlite deposits around the world. Most notably I would like to thank Barbara Scott Smith for allowing me to study her vast kimberlite collections, learn from her veteran experience as an kimberlite geologist and industry consultant, and for her valuable feedback regarding my research.   Finally, I would like to thank the SEM and EMP lab staff, Mati, Edith, Elizabetta, Jenny, and Lan, for their guidance using analytical equipment and for their ever-friendly and accommodating nature.  Introduction - Statement of the Research Problem  1  1. Introduction 1.1 Statement of the Research Problem Our understanding of the near-surface emplacement-related processes that kimberlite magmas undergo are relatively poor in comparison to other volcanic systems. Modern examples of basaltic and rhyolitic eruptions allow for direct observation of processes such as volcanic fragmentation, degassing, and pyroclastic fall or flow deposition. These observations allow for a meaningful interpretation of the textures and mineralogy of the resulting rock types as encountered in the field. Kimberlite magmatism however, was generally confined to the Proterozoic and Phanerozoic with the notable exception of the Quaternary Igwisi Hills volcanic field in Tanzania (Kjarsgaard, 2007; Brown et al., 2012; Willcox et al., 2015). There are no modern observable kimberlite eruptions upon which to base our interpretation of the eruption process.  It is also challenging to draw parallels between the volcanic processes in these well-established basaltic and rhyolitic pyroclastic systems to those envisioned for kimberlitic eruptions. While some types of pyroclastic kimberlites referred to as Fort-à-la-Corne-type pyroclastic kimberlites display evidence of pyroclastic fall deposition including clast supported textures, pyroclast sorting (figure 1.1A), and fine-particle elutriation (figure 1.1B), other types of pyroclastic kimberlite referred to as Kimberley-type pyroclastic kimberlites do not (Hetman et al., 2004; Scott Smith, 2008a). These Kimberley-type pyroclastic kimberlites are characterized by the presence of abundant spherical magmaclasts which contain either an angular pyroclast or round pyrocryst enveloped by a generally thin melt selvage (figure 1.1C); these pyroclasts appear to “float” within a matrix dominated by microlitic clinopyroxene and interstitial serpentine (figure 1.1D). These deposits also display continuous gradational textural transitions between pyroclastic and coherent textures (figure 1.1C). Kimberlite geologists have proposed that these features demonstrate that Kimberley-type pyroclastic kimberlites may result from continuous in-situ subvolcanic  Introduction - Statement of the Research Problem  2   Figure 1.1 – Textures of Fort-à-la-Corne type pyroclastic kimberlites and Kimberley-type pyroclastic kimberlites. Fort-à-la-Corne-type pyroclastic kimberlites display evidence of emplacement by conventional processes such as those envisioned for pyroclastic basaltic eruptions including (A & B) clast-supported textures, (A) sorting of pyroclasts, and (B) fine-particle elutriation. Kimberley-type pyroclastic kimberlites by contrast do not display evidence of sorting or significant fine-particle elutriation. They display (C & D) clast supported textures with abundant spherical magmaclasts hosted within a (D) fibrous microlitic interclast matrix dominated by clinopyroxene with interstitial serpentine. Figures A & B are modified after Scott Smith, 2008b and figures C & D are modified after Hetman et al., 2004. degassing and fragmentation processes causing textural modification of a rapidly cooling subsurface magma column (Clement, 1982; Field & Scott Smith, 1999; Hetman et al., 2004; Skinner, 2008; Scott Smith, 2008a). This model attributes the matrix-supported textures to the precipitation of interclast clinopyroxene and serpentine from exsolved magmatic volatiles in a high-pressure fluidized environment during intrusive explosion. Volcanology experts however contest that these features are more likely to be the result of low-grade metamorphism overprinting a partially welded pyroclastic deposit formed by more conventional processes such as those envisioned for basalts (Sparks et al., 2006; Walters et al., 2006; Introduction - Kimberlite Defined  3  Stripp et al., 2006; Sparks et al., 2009; Buse et al., 2010). Implicit in this difference of interpretation is to what extent the textures and mineralogy of kimberlites can be used to interpret the emplacement processes of these magmas, and to what extent those features reflect the effects of hydrothermal alteration.  The goal of this research is to describe in great detail the petrography of pyroclastic, transitional pyroclastic and coherent kimberlites through continuous drill-core intersections from present day surface down to 200 m depth and throughout an entire pipe using multiple drill-cores, with the aim of constraining the processes related to their formation. The focus of the research is on the interclast matrix assemblages and crustal xenolith populations, their textures and compositions, as a means of resolving the difference of opinion in the current models. This research will use samples from four complete drill-hole intersections through the Renard 65 kimberlite pipe owned by Stornoway Diamond Corporation, one of four economic pipes included in the Renard Mine. By characterizing the nature of in-situ magmatic processes involving kimberlite magma and silicic crustal xenoliths during emplacement, and subsolidus processes involving exsolved deuteric fluids interacting with and modifying the kimberlite both during and immediately following emplacement, I provide the necessary constraints to propose an improved model for the formation of Kimberley-type pyroclastic kimberlites. These criteria are also applied to improve the identification of individual batches of kimberlite magma in Renard 65 allowing for improved criteria with which to constrain the distribution of economic value within the pipe.  1.2 Kimberlite Defined In 1869 on the Dutoitspan and Bultfontein farms in South Africa, round, muddy, water-filled depressions known as “pans” were discovered to be underlain by “blue rock”, a rock containing diamonds; it would later come to be known that these “pans” were the weathered surface expressions of kimberlite pipes (Mitchell, 1985). The discovery of these diamond bearing pans led to the development of four Introduction - Kimberlite Defined  4  artisanal diamond mines around which the town of Kimberley was founded. The “blue rock” thus inherited the name, Kimberlite. Today, much more is known about the nature of kimberlite: its origin, its journey through the mantle and crust, and the volcanological aspects of its emplacement. As a result, the definition of kimberlite has become much more precise, evolving with contributions from many authors (Skinner and Clement, 1979; Dawson, 1980; Clement et al., 1984; Smith et al., 1985; Mitchell, 1979, 1986, 1994, 1995, and more). In recognition of both the hybrid nature of kimberlites, and the number of petrologically similar, but genetically different rock types including orangeite, lamproite, melnoite, alnoite, and olivine melilitite. The IUGS Subcomission on the Systematics of Igneous Rocks provides the most current definition of kimberlite as follows (Wooley et al., 1995). Kimberlites are defined as “volatile-rich (dominantly CO2) potassic ultrabasic rocks commonly exhibiting a distinctive inequigranular texture resulting from the presence of macrocrysts (a general term for large crystals, typically 0.5 – 10 mm in diameter) and, in some cases, megacrysts (larger crystals, typically 1 – 20 cm) set in a fine-grained matrix. The assemblage of macrocrysts and megacrysts, at least some of which are xenocrystic, includes anhedral crystals of olivine, magnesian ilmenite, pyrope, diopside (in some cases subcalcic), phlogopite, enstatite and Ti-poor chromite. Olivine macrocrysts are a characteristic and dominant constituent in all but fractionated kimberlites. The matrix contains a second generation of primary euhedral to subhedral olivine, which occurs together with one or more of the following primary minerals: monticellite, phlogopite, perovskite, spinel (magnesian ulvöspinel – magnesiochromite – ulvöspinel – magnetite solid solutions), apatite, carbonate and serpentine. Many kimberlites contain late-stage poikilitic mica belonging to the barian phlogopite-kinoshitalite series. Nickeliferous sulfides and rutile are common accessory minerals. The replacement of earlier formed olivine, phlogopite, monticellite and apatite by deuteric serpentine and calcite is common. Evolved members of the group may be poor in or devoid of, macrocrysts, and composed essentially of second Introduction - Kimberlite Classified  5  generation olivine, calcite, serpentine, and magnetite, together with minor phlogopite, apatite and perovskite” (Wooley et al., 1995). 1.3 Kimberlite Classified In 1887 kimberlite was given its first petrographic name, “porphyritic mica-bearing peridotite, a volcanic breccia” by the British Association for the Advancement of Science in Manchester (Lewis, 1887). Percy Wagner (1914) proposed the first kimberlite classification scheme in which he recognized two types of diamond-bearing rock, one which he termed basaltic kimberlite (kimberlite) and one which he termed micaceous kimberlite (orangeite). This classification scheme held for half a century before the next major revision to kimberlite nomenclature and classification by Skinner and Clement (1979) and Clement et al., (1984). Skinner and Clement proposed that the most important component for kimberlite classification was the groundmass modal mineralogy, and proposed that kimberlites be named on the basis of their most abundant groundmass mineral. They noted that separating xenocrystic and phenocrystic olivine was challenging if not impossible at times and therefore not useful in the classification of kimberlites. The development of this classification scheme was crucial, as it paved the way for a standardized means of comparing kimberlites from all different localities. It resulted in the discovery of kimberlite-related rock types, and provided the earliest means of differentiating between the solidified products of different batches of kimberlite melt within a pipe (or phases of kimberlite in a geological model). Over the decades that followed, the terminology and classification schemes of kimberlites evolved in tandem with, but in some ways independent from mainstream volcanology, owing in large part to the economic evaluation of kimberlite deposits by diamond mining companies. Economic kimberlite geologists emphasized the use of size and abundance descriptors that were of practical use for the ease of communication in the evaluation of diamond deposits, and nomenclature that reflects the features in pyroclastic kimberlites that are interpreted to have formed by unconventional pyroclastic processes. Volcanologists emphasized the use of terminology and nomenclature schemes that were consistent with mainstream volcanology.   Introduction - Kimberlite Classified  6  A significant number of papers were published on kimberlite classification schemes, terminology, and nomenclature (Dawson, 1971, 1980; Hawthorne, 1975; Clement and Skinner, 1985; Mitchell, 1986, 1995; McPhie et al., 1993; Field and Scott Smith, 1998; Cas et al., 2008a, 2009; Scott Smith et al., 2013). Despite some disagreement about the use of terminology, and size and abundance schemes, all of these papers agreed on the need for descriptive terminology that does not contain genetic connotations. The Kimberlite Classification and Terminology framework presented by Scott Smith et al. (2013) is a collaboration from a number of accredited researchers with both economic geology and volcanology backgrounds, and builds upon this half century of evolving schemes. According to Scott Smith et al. (2013) all kimberlites may be subdivided into three major components: (i) compound clasts, (ii) crystals, and (iii) interstitial matrix (figure 1.2). If these components are carefully documented and described, they may provide sufficient evidence for the identification of sub-components (figure 1.2). As a rather simple example, crystals described as “olivine with euhedral crystal shapes” may be reclassified as olivine phenocrysts. For a more complex example, compound clasts in a kimberlite rock sample that are described as “angular or subangular clasts composed of coherent kimberlite, with broken olivine macrocrysts truncated by the clast margins” can be reclassified as autoliths; this on the basis that they are interpreted to be clasts of previously solidified kimberlite entrained within a subsequent kimberlite emplacement event.  Introduction - Kimberlite Classified  7   Figure 1.2 – Conceptual framework for the description of kimberlite components (Scott Smith et al., 2013). Three classes of kimberlite components are proposed (A) Compound Clasts, (B) Crystals, and (C) Interstitial Matrix. Increasing subdivision of components within each class requires increasing supporting evidence of the processes involved in their formation. In addition to the kimberlite component framework, Scott Smith et al. (2013) propose the use of practical kimberlite-specific size and abundance nomenclature for kimberlite components (figure 1.3). Unlike size and abundance descriptors for pyroclastic rocks, these are formulated around the commonly observed ranges of size and abundance for kimberlite components. This research program involves the frequent use of volcanological and kimberlite specific terminology that may be unfamiliar to the reader; a glossary of kimberlite and related terminology is provided using definitions from Kimberlite Terminology and Classification (Scott Smith et al., 2013) and from A Glossary of Kimberlite and Related Terms (Scott Smith et al., in press.).   Introduction - Kimberlite Classified  8   Figure 1.3 – Size and abundance descriptors for compound clasts and crystals (Scott Smith et al., 2013). Coarser materials including lithic clasts and xenoliths are sized on a different scheme than finer materials including crystals, magmaclasts and accretionary clasts. For lithic clasts and xenoliths, the descriptor [xenolith] may be substituted with [autoclast], [autoliths] or [epiclast] where appropriate. Similarly, for crystals, magmaclasts and accretionary clasts, the descriptor [crystal] should be substituted with [magmaclast], [accretionary clast] or the crystal type (i.e. [olivine]). The use of this size and abundance nomenclature does not require any genetic interpretation, and provides ease of communication of kimberlite components by using kimberlite appropriate ranges.Introduction - Kimberlite Classified  9  The application of the kimberlite component framework and size/abundance descriptors of Scott Smith et al. (2013) culminates in a 5-step framework for the description, classification and interpretation of kimberlites (figure 1.4). Each stage involves an increasing degree of interpretation, requiring sufficient megascopic, macroscopic and microscopic textural and mineralogical evidence in support of textural-genetic classifications, construction of volcanic/intrusive spatial context, and interpretations about emplacement processes. Stage 1 begins with general rock descriptions relating to alteration, structure, texture, and the identification of kimberlite components, and is generally based on megascopic and macroscopic observations. Rock descriptions do not require the identification of the parental magma type (kimberlite, lamproite, melnoite, alnoite, olivine melilitite), and example rock names include ‘uniform, xenolith-poor, medium-grained, olivine macrocryst-rich rock’. During this description stage, the use of non-genetic terminology is essential, as interpretations about parental magma type, or genetic processes related to emplacement may change over the course of characterizing a deposit, but these changes in interpretation will have no bearing or consequence on the original rock descriptions.  Stage 2 involves the petrographic characterization of groundmass mineral assemblages, the determination of the parental magma type based on groundmass modal mineralogy, and the subsequent (non-genetic) mineralogical classification. The determination of parental magma type allows for the classification of the rock as a kimberlite or a related rock type, a determination with both economic and genetic significance. The mineralogical classification provides direct evidence of groundmass modal mineralogy, and allows for the potential to differentiate between separate batches of solidified kimberlite magma (or phases of kimberlite).  Introduction - Kimberlite Classified  10    Figure 1.4 – 5 step systematic framework for the description, classification, and subsequent interpretation of kimberlites (Scott Smith et al., 2013). (Stage 1) Using descriptive terminology and kimberlite specific size/abundance descriptors, rocks are described and named (i.e. medium-grained olivine macrocryst-rich rock). (Stage 2)  Introduction - Kimberlite Classified 11  Stage 3a marks the beginning of the interpretation process, and any changes made prior to this stage will not impact the descriptions in the previous two stages. It is at this stage appropriate to use important features such as texture and mineralogy of component features to distinguish between coherent and volcaniclastic kimberlites. Coherent kimberlites are characterized by an interstitial matrix of microphenocrysts, groundmass, and mesostasis mineral assemblages (figure 3.1C) (Scott Smith et al., 2013). Volcaniclastic kimberlites are defined as kimberlites containing a large proportion of volcanic particles which may be any shape or size, with no implicit connotations related to clast-forming processes, transport or depositional processes, or environment (Scott Smith et al., 2013).  Stage 3b advances the textural-genetic interpretation process by differentiating between sub-classes in coherent and volcaniclastic kimberlites. For coherent kimberlites, differentiating between intrusive coherent kimberlite (hypabyssal kimberlite) and extrusive coherent kimberlite (lavas and flows) requires megascopic context and detailed knowledge of concordant or discordant relationships to the host rock. Volcaniclastic kimberlites can be differentiated into (i) pyroclastic kimberlites, (ii) resedimented kimberlites and (iii) epiclastic kimberlites. Pyroclastic kimberlites are deposited by primary pyroclastic processes during explosive volcanic eruptions, and do not show evidence of active sedimentary processes (Scott Smith et al., 2013). Resedimented volcaniclastic kimberlites are the product of re-sedimentation of unconsolidated pyroclastic material, along with extraneous surface materials (Scott Smith et al., 2013); the identification of RVK’s is largely dependent on the macroscopic identification of sedimentary textures (bedding, sorting) but these textures are reproducible by pyroclastic processes as well, so the identification of surface materials within the kimberlite (i.e. tree bark) is generally required to determine whether a kimberlite is an RVK with a significant degree of confidence. Epiclastic kimberlites are secondary sedimentary deposits containing epiclasts derived from exposed and eroded kimberlite (Scott Smith et al., 2013). As a final step in this stage, pyroclastic kimberlites may be differentiated as either Kimberley-type pyroclastic kimberlites (KPK) or Fort a la Corne-type pyroclastic kimberlites (FPK). Kimberley-type Introduction - Kimberlite Classified 12  pyroclastic kimberlites are “a class of pyroclastic kimberlite that encompass a variety of textural rock types with unifying textural and component characteristics that are distinctly different from FPK’s; they are generally massive, unsorted, matrix- to clast-supported with abundant xenoliths, magmaclasts, olivine macrocrysts, and less common autoliths, set in a fine-grained interclast matrix dominated by microlitic diopside and phlogopite, and irregularly distributed primary groundmass minerals” (Scott Smith et al., in press). Fort-à-la-Corne-type pyroclastic kimberlites are “a class of pyroclastic kimberlite that also encompasses a variety of textural rock types with unifying textural and component characteristics, which are in many cases comparable to basaltic pyroclastic rocks; FPK’s are generally dominated by melt-bearing pyroclasts and liberated pyrocrysts (melt-free crystal pyroclasts) in any relative proportions, with a clast-supported texture and variably massive to graded and bedded texture” (Scott Smith et al., in press).  Stage 4 combines detailed rock descriptions and textural-genetic interpretations with megascopic spatial context and morphological distribution (generally requiring detailed drilling and mapping of kimberlite distribution) to allow for the identification of pipes, dykes, sills, lava flows, extra-crater deposits, and epiclastic sedimentary deposits. Stage 5 is the final and most interpretive aspect of kimberlite description and classification. It involves interpretation of the emplacement processes by which a kimberlite was formed. Examples of process interpretations include (effusive lava flows, pyroclastic spatter or fallout, fluidisation, among many others). These genetic process interpretations can be added to the rock name derived through stages 1 – 4 (i.e. clast-supported, very xenolith-rich RVK mass flow deposit). The presentation of the petrography of the Renard 65 kimberlites in this research generally follows this five-step framework with mineralogical and textural descriptions of kimberlite components accompanied by mineral chemistry (Stage 1) presented in Sect. 3.1 Geological Unit Summaries, and Sect. 3.2 Mineral Chemistry. The Renard 65 kimberlites are then assigned a parental magma type and Introduction - Kimberlite Emplacement 13  mineralogical classification (Stage 2), and textural-genetic classifications (Stage 3) in Sect. 4.0 Renard 65 Rock Classification. The discussion that follows will focus on the results of the previous sections in interpreting the emplacement and post-emplacement-processes that affected the Renard 65 kimberlites (and perhaps also similar kimberlites at other localities), and the identification of separate phases of kimberlite (Stage 4 & 5). 1.4 Kimberlite Emplacement The classic model of kimberlite magmatism was largely based on detailed studies of South African kimberlites (Mitchell, 1986, 1995; Mannard, 1962; Dawson, 1971; Hawthorne, 1975; Clement, 1982) which are analogous to Kimberley-type pyroclastic kimberlites (Scott Smith et al., 2013). In this classic model (figure 1.5), a kimberlite pipe is depicted as a carrot-shaped body with an upper crater facies (containing solidified lavas, pyroclastic material, and resedimented volcanics), a conical diatreme facies (containing volcaniclastic kimberlites with abundant magmaclasts, formerly referred to as tuffisitic kimberlites with pelletal lapilli) and a deep and irregular root zone (containing coherent kimberlites). Figure 1.5 – Classic model of an idealized kimberlite magmatic system (Mitchell, 1984, 1995). This early kimberlite pipe model based on South African occurrences depicts a steep-sided pipe with upper crater facies, middle diatreme facies, and lower root zone. The pipe is shown to both cross-cut and be cross-cut by precursor and late magmatic hypabyssal kimberlite dykes and sills.  Introduction - Kimberlite Emplacement 14  Emplacement of the kimberlite pipe is pre-dated by the intrusion of precursor dykes and sills, and post-emplacement dykes and sills may cross-cut the pipe as well. Clement and Skinner (1979, 1985) and Clement (1982) described numerous South African kimberlites and observed that while hypabyssal facies (coherent kimberlites) kimberlites were by and large consistent in their mineralogy and textures across localities, the diatreme facies kimberlites were much more variable in their appearance, but they displayed a repeated range of textures and mineralogy. The authors concluded that these diatreme facies kimberlites must have resulted from a gradational modification of the underlying kimberlite magma erupting from the root zone by near surface emplacement processes.  Based on detailed studies of these South African kimberlites, two modern theories have emerged to describe the emplacement process that could produce diatreme facies, matrix-supported volcaniclastic kimberlite, with abundant xenoliths, xenocrysts and magmaclasts, in a fine-grained microlitic matrix. The first emplacement model envisions that kimberlite eruptions are characterized by subvolcanic fragmentation, degassing in response to decompression-induced exsolution of magmatic volatiles, and rapid cooling of a magma column (Dawson, 1971; McGetchin, 1968; Woolsey et al., 1973, 1975; Clement, 1975; McCallum, 1976, 1985; Clement & Skinner, 1979, 1985; Clement, 1982; Clement & Reid, 1989; Field and Scott Smith, 1999; Hetman et al., 2004; Skinner 2008; Scott Smith, 2008); the rapid ascent of exsolved magmatic volatiles to surface during emplacement is argued to result in a state of fluidization resulting in fragmentation of the melt into melt-droplets which cool to form magmaclasts, and quenching of the exsolved volatiles to produce the microlitic interclast matrix assemblage. The second emplacement model envisions a kimberlite diatreme excavated during explosive volcanic fragmentation and subsequently infilled by mixtures of lithic wall-rock clasts from rock-bursts and slumping, and pyroclastic material deposited aerially from the collapsing eruption column. This clastic material is then altered by the introduction of hydrous meteoric fluids resulting in the in-situ recrystallization of the matrix to a microlitic Introduction - Regional and Local Geology of the Renard Kimberlites 15  interclast matrix assemblage (Stripp et al., 2006; Sparks et al., 2006; Walters et al., 2006; Hayman et al., 2008, 2009; Buse & Sparks, 2010; Gernon et al., 2012).  A more detailed discussion of these emplacement models, and their viability in accounting for the features of the Renard 65 kimberlites and similar Kimberley-type pyroclastic kimberlites at other localities is presented in Sect. 5.2, Kimberlite Emplacement: Processes of Diatreme Formation and Infill. 1.5 Regional and Local Geology of the Renard Kimberlites The Renard kimberlites are part of the Otish kimberlite volcanic event, which includes 12 known kimberlite pipes emplaced between 550 +/- 3.5 Ma (Girard, 2001; Moorhead et al., 2002) and 640.5 +/- 28 Ma (Birkett et al., 2003) within the Opinaca and Opatica subprovinces of the Superior province, Canada (Figure 1.6). The Superior province is bounded by Paleoproterozoic provinces to the west, north and east, Figure 1.6 – Geology of the Superior Province (Percival, 2007). The Renard kimberlites are located in the North-Eastern portion of the Opatica (OcS) and Opinaca (OnS) subprovinces at 52°49’ N and 72°12’ W. Introduction - Regional and Local Geology of the Renard Kimberlites 16  and the Mesoproterozoic Grenville Province to the southeast. Current views on the amalgamation of the Superior province suggest that a number of small Mesoarchean continental fragments, and Neoarchean oceanic plates were aggregated between 2.72 and 2.68 Ma, and subsequently affected by post-orogenic events. The Opatica subprovince (OcS) consists of primarily metaplutonic tonalities (2.82 Ga), tonalite-granodiorite (2.70 – 2.77 Ga), granite and pegmatite (2.68 Ga) (Percival, 2007). The Opatica subprovince exhibits polyphase deformation with west-vergent shear zones overprinted by south-vergent structures (Sawyer and Benn, 1993). The Opinaca subprovince consists of metagreywacke, migmatite, and granite with polydeformed schists along the belt margins of the subprovince. The inner portions of the subprovince are metamorphosed to amphibolite and granulite facies (Percival, 2007).  Figure 1.7 – Geological setting of the Renard cluster (after Muntener and Scott Smith, 2013). (A) The Renard cluster is part of the Otish kimberlite field located in the eastern Superior province. (B) There are 9 kimberlite pipes in the Renard cluster; Renard 65 is the largest of the 9 kimberlite pipes. Introduction - Regional and Local Geology of the Renard Kimberlites 17  The Renard property, owned by Stornoway Diamond Corporation includes a field of 9 diamondiferous kimberlite pipes (figure 1.7B), and a number of hypabyssal dykes with an age of 640.5 +/- 28 Ma determined by 206Pb/238U ratios of groundmass perovskite (Birkett et al., 2003). Previous studies on the Renard bodies, including a detailed economic evaluation of Renard 3 by Muntener and Scott Smith (2013), have shown that the Renard kimberlites display features that are consistent with the Kimberley-type pyroclastic kimberlites. The current surface expressions of the pipes are believed to be exposing the lower diatreme to root zone facies (Birkett et al., 2003). The geological model of the Renard 65 pipe developed by Stornoway Diamond Corporation (figure 1.8) includes three main pipe infilling geological units of coherent to volcaniclastic kimberlite (Kimb65a, Kimb65b and Kimb65d), one unit consisting of hypabyssal dykes and sills that occur throughout the pipe (Kimb65c), and two fractured and brecciated country rock units that envelope the kimberlite units (CRB and CCR).  Introduction - Regional and Local Geology of the Renard Kimberlites 18   Figure 1.8 – Geological Model of the Renard 65 Pipe. (A) Renard 65 geological model showing only kimberlite units, with locations of drill-holes used in this study. (B) Renard 65 geological model showing both kimberlite and non-kimberlite rock-types. (C) Renard 65 geological model in plan view (Bagnell et al., 2013). Introduction - Research Objectives 19  1.6 Research Objectives This research on the Renard 65 kimberlites provides a unique opportunity to examine pipe-infilling kimberlite rock types in a spatially dense and representative manner. As a consequence, the objectives of this research are wide-ranging and can be divided into those relevant to economic geology and mining of Renard 65 (points 1 – 3) and those contributing to fundamental volcanology and petrology (points 4 – 5). 1) Mineralogical and textural classification of the Renard 65 kimberlites. 2) Assignment of the pipe-infilling rock-types to geological units by isolation of sets of rock types with consistent mineralogy and kimberlite component features that are distinctly different from adjacent rock types. 3) To determine from these geological units, the number of separate magmatic events or ‘phases of kimberlite’ in which emplacement-related processes may account for any variability in kimberlite component features.  4) To use the mineralogy, textures, and megascopic features of the Renard 65 kimberlites and other similar kimberlites at different localities to critique first the emplacement models and models on the origin of the interclast matrix in Kimberley-type pyroclastic kimberlites. 5) To determine the role of magmatic and subsolidus interactions between silicic crustal xenoliths and their host kimberlite and to model the implications of xenolith assimilation on the mineralogy and the behavior of magmatic volatiles during emplacement.  Methods - Core Logging and Polished Slab Descriptions 20  2. Methods 2.1 Core Logging and Polished Slab Descriptions Core logs were completed for four drill holes (R6-01, R65-31, R65-33, and R65-34) totalling over 700 m in length, to establish baseline megascopic and macroscopic descriptions of the characteristics of all geological units (Appendix A); full core photographs of an additional 22 drill-holes were examined to ensure a spatially representative nature of geological units contained in the four drill holes used in this study. The core logs (Appendix A) are written as a sequence of intervals and sub-intervals. Each interval spans a measurable range of drill core in which the rock characteristics are consistent and uniform. Intervals separate different rock types with distinct differences in kimberlite components (i.e. rock texture, or olivine macrocryst size, etc.) in drill core, and sharp contacts are present between intervals. One interval may be broken into a set of sub-intervals to denote lesser changes in components within the same geological unit (i.e. changes in country rock xenolith dilution, or gradational changes in kimberlite texture). Each interval description begins with a summary statement on the macroscopic colour, texture and geological unit assignment. This is then followed by individual descriptions of kimberlite components: CR xenoliths, olivine macrocrysts and phenocrysts, magmaclasts, groundmass mineral assemblages, interclast matrix mineral assemblages, and any unique features (i.e. veining, flow sorting, etc.). The description is concluded with a general comment on the degree of confidence in the geological unit assignment, highlighting factors affecting the decision.  While full drill core is essential for providing a megascopic perspective of the geology, the rough and rounded surfaces of the drill core make it challenging to observe finer macroscopic features such as details of kimberlite rock texture. In order to improve the macroscopic resolution of drill-core for more detailed description, representative drill core samples for all intervals and sub-intervals were prepared into polished slabs. A total of 43 drill core samples were selected from the four drill holes, of which 27 were prepared into polished slabs; in the preparation of a polished-slab, a drill core sample is dry cut Methods - Kimberlite Petrography and Scanning Electron Microscopy 21  lengthwise, with one half of the drill core finely polished on the inner flat surface using water-free polishing techniques. Rock descriptions for each polished slab under binocular microscope were completed in the format of a very detailed single interval description in a core log (Appendix B). The combined core logs and polished slab descriptions form the basis of the geological unit summary (figure 3.1). They also provide a context for the addition of microscopic observations of kimberlite components, in a representative manner, using thin-section petrography and scanning electron microscopy. 2.2 Kimberlite Petrography and Scanning Electron Microscopy Microscopic kimberlite components that are not readily observed in drill core or polished slab (i.e. groundmass and interclast matrix mineral assemblages and xenolith reaction mineral assemblages), were characterized using thin-section petrography and scanning electron microscopy. From the offcut portion of the polished slabs, a total of 35 polished thin-section locations were selected to capture microscopic kimberlite components observed in the polished slab. These thin-sections were described using thin-section petrography, with more detailed textural description and mineral identification using a PHILLIPS XL30 scanning electron microscope (SEM) energy dispersive X-ray spectrometer (EDS) with Bruker Quantax 200 Microanalysis system and light element XFLASH 6010 detector at the Earth, Ocean and Atmospheric Sciences Department, at the University of British Columbia. EDS was primarily used for quantitative mineral identification, and X-ray composition element mapping. In order to reliably determine groundmass modal mineralogy for each geological unit, and to assess the consistency of that modal mineralogy spatially throughout the unit, element distribution maps were routinely produced for 1000 x 1000 µm areas of kimberlite groundmass, in 3 – 4 areas per thin section, and with thin sections at 5 – 6 depths within the geological unit. This approach was found to be viable, given the extremely small size of the groundmass mineral assemblages, and the observed consistency in modal mineralogy both Methods - Mineral Chemistry 22  within different areas of the same sample, and different depths within the same geological unit. The SEM was also used to document textures of groundmass, interclast matrix, and xenolith reaction mineral assemblages, and to improve mineral identification. The petrographic and SEM investigation of microscopic kimberlite components in each geological unit were combined with megascopic observations from core logs, and macroscopic observations from polished slab descriptions, to complete a comprehensive geological unit summary table, presented in Sect. 3.1 Geological Unit Summaries.  2.3 Mineral Chemistry Quantitative mineral chemistry data was collected for xenolith reaction mineral assemblages, groundmass mineral assemblages, and interclast matrix mineral assemblages in each geological unit. The mineral chemistry is separated by both geological unit and textural position as presented in Sect. 3.2 Mineral Chemistry. Mineral chemistry results are presented as representative averages; for non-averaged tables please refer to Appendix C – Renard 65 Mineral Chemistry. Electron-probe micro-analyses of all minerals were completed on a fully automated CAMECA SX-50 instrument, operating in the wavelength-dispersion mode, at the Earth, Ocean and Atmospheric Sciences Department, at the University of British Columbia. All analyses were completed at an excitation voltage of 15 kV, 20 s peak count times, and 10 s background count times, with the following peak and background count time exceptions: Sr in carbonates increased to 40 s/20 s, Cl and Sr in phlogopite increased to 40 s/20 s for Cl and 80 s/40 s for Sr, K in pyroxenes increased to 40s/20s, and Ni in spinel increased to 40s/20s. The following spot diameters were used: 10 µm for carbonates, apatite and phlogopite, 5 µm for pyroxenes, olivine and perovskite, and 2 µm for spinels. The excitation voltage for amphiboles was increased to 20 kV. All data reduction was done using the 'PAP' (Z) method (Pouchou & Pichoir 1985). For carbonates, O was determined by stoichiometry and C was determined by difference. Results - Geological Unit Summaries 23  3. Results 3.1 Geological Unit Summaries The geological units Kimb65a, Kimb65b, Kimb65c, Kimb65d, CCR and CRB characterize rock types encountered in drill-core with a specific range of textures and component features that are repeated over large intervals. Kimberlite component features include (i) size, shape, abundance and reaction of country rock xenoliths, (ii)size, crystal shape, abundance and replacement of olivine macrocrysts and microcrysts, (iii)primary kimberlite groundmass mineral assemblage and textures and (iv) interclast matrix mineral assemblage and texture. A summary table of the features of kimberlite components for the kimberlite geological units is presented in Table 3.1. The description of these features in the following geological unit summaries is based on the combined results from macroscopic drill-core logging, binocular microscope descriptions of polished drill core slabs, petrographic thin-section descriptions, and detailed textural examination on an SEM using BSE imaging and EDS spot analyses and element distribution mapping. In this section, detailed representative geological unit summaries are presented for: (1) Kimb65a: xenolith-dominated olivine-poor magmaclast-rich volcaniclastic kimberlite, (2) Kimb65b: xenolith poor olivine-rich magmaclast-poor coherent to lesser volcaniclastic kimberlite, (3) Kimb65c: xenolith-poor olivine-rich macrocrystic coherent kimberlite, and (4) Kimb65d: xenolith-rich olivine-poor magmaclast-rich volcaniclastic kimberlite. For the purposes of this research program, the fractured and brecciated country-rock geological units identified by Stornoway Diamond Corporation are not examined in any significant detail; though very basic descriptions of these units are included at the end of this chapter.     Results - Geological Unit Summaries 24  Table 3.1 - Renard 65 Geological Unit Summary   Geological Unit Summaries - Kimb65a 25  3.1.1 Kimb65a The following description is a representative summary of one or more rock types with a consistent set of kimberlite features, repeated over large continuous intersections within two vertical drill-holes (R65-31 & R65-33) reaching 191 m and 200 m depth (figure 1.8). Visual examination of a further 4,000 m of drill core from 23 additional vertical and sub-vertical drill holes confirms that rocks consistent with the characteristics of Kimb65a form a vertical-conical body occupying approximately 65 % of the pipe by volume (figure 1.8).  Kimb65a can be described as a xenolith-dominated olivine-poor magmaclast-rich volcaniclastic kimberlite. These rocks display an inequigranular and volcaniclastic texture with high dilution of relatively mildly-reacted angular country rock xenoliths, pseudomorphed olivine macrocrysts, and a magmaclastic texture defined by (i) cored, (ii) uncored, and (iii) incipient magmaclasts, all hosted within an interstitial matrix composed of microlitic phlogopite and diopside, and irregularly distributed primary kimberlite groundmass minerals (figure 3.1). The country rock xenoliths (CRx) are micro to small macroxenoliths (0.5 mm pulverized fragments to <8 cm clasts) with angular to subangular clast shapes. Modal abundances are generally very consistent (±5 %) over large intervals (meters to tens of meters), and range from 40 – 90 %. Xenolith composition varies from granitoid to gneiss protoliths with well preserved igneous and metamorphic textures (figure 3.1A). The crustal xenoliths consist of quartz, plagioclase, alkali feldspar, and biotite; the are commonly partially recrystallized to amphibole at the grain boundaries between quartz and feldspars (figure 3.1E), while biotite commonly shows variable chloritization with intergrowths of pyrite forming elongated crystals between the original cleavage planes of the biotite (figure 3.1F). Xenoliths are either hosted as discrete clasts/xenocrysts within the interclast matrix (figure 3.1A), or form the cores of magmaclasts (figure 3.1B).  Geological Unit Summaries - Kimb65a 26   Figure 3.1 - (A) Polished slab (PS) with volcaniclastic texture containing visibly distinct subround magmaclasts with calcite pseudomorphed olivine macrocrysts. (B) Variably altered cored and uncored magmaclasts consisting of dark brown kimberlite groundmass altered to a pale green-blue colour; the outlines of the former magmaclast are still visible. (C) Transitional volcaniclastic texture with incipient magmaclasts forming irregular mottled patches of coherent kimberlite groundmass (dark brown) separated by interclast matrix (light beige). (D) PPL image of uncored and cored magmaclasts; note that the margins of the magmaclasts show evidence of zoning, in which calcite pseudomorphs of olivine are overprinted by chlorite and serpentine, and perovskite is altered. (E) BSE image showing the occurrence of amphibole in crustal xenoliths crystallizing at the interfaces between quartz and feldspar; amphiboles are partially altered to chlorite. (F) BSE image showing a fully chloritized biotite crystal in crustal xenoliths with intergrowths of elongated pyrite crystals conforming to the cleavage planes of the biotite. Geological Unit Summaries - Kimb65a 27  Magmaclasts (mc) range in modal abundance from 10 – 20 %, from a minimum size of 1 mm up to a maximum size of 12 cm, forming round to subround clasts. Clasts may be entirely composed of solidified kimberlite magma (figure 3.1A & 3.1D), or they may form whole or partial selvages of variable thickness on olivine macrocrysts and country rock xenoliths (figure 3.1B). In some samples, magmaclasts are less spherical and visibly distinct, occurring as partially formed clasts with an interconnected texture; this is referred to as ‘transitional texture’ between coherent and volcaniclastic (figure 3.1C); these are here referred to as incipient magmaclasts. In other samples, magmaclasts show evidence of zoning (figure 3.1B & 3.1D). In thin-section, zoning is characterized by (i) increasing degrees of serpentinization of olivine (figure 3.1D), and (ii) alteration of perovskite to Mn-ilmenite + pyrite + calcite (figure 3.1D, 3.2A & 3.5E). Olivine (ol) macrocrysts range in modal abundance from 5 – 15 %, forming anhedral, round and occasionally sub-angular crystal shapes, ranging in size from 1 – 4 mm and rarely as large as 6 mm. Olivine phenocrysts account for 3 – 12% modal abundance of the rock, forming subhedral to euhedral crystal shapes ranging in size from 0.25 – 1 mm. In general, modal abundances of both olivine macrocrysts and microcrysts decrease with increasing modal abundance of country-rock xenoliths. In Kimb65a, primary olivine is not preserved either within magmaclasts or as discrete crystals within the interclast matrix. These grains are identified as olivine on the basis of their size, shape and modal abundance, but are fully pseudomorphed by calcite, and variably overprinted by serpentine and chlorite. The relative abundance of these minerals pseudomorphing after olivine depends on the textural position of the olivine crystal. Olivines hosted closer to the center of larger uncored magmaclasts show very low degrees of serpentine alteration (figure 3.1A, 3.1D & 3.2D), while those along the margins of the magmaclasts (figure 3.1D) or hosted as discrete crystals in the interclast matrix (figure 3.2C) show increasing degrees of serpentine overprinting; serpentine is variably altered to chlorite (figure 3.2C).  Geological Unit Summaries - Kimb65a 28   Figure 3.2 – (A) BSE image of a zoned uncored visibly distinct magmaclast, showing compositional zonation between the inner and outer portions of the magmaclast; in the margins of the clast, perovskite is replaced and partially resorbed, and calcite after olivine pseudomorphs are strongly overprinted by serpentine and chlorite. (B) Perovskite (pv) abundance within core of magmaclast (A) illustrated using Ti element-distribution map. (C) Olivine macrocryst with partial selvage of kimberlite groundmass; olivine is pseudomorphed by serpentine, and overprinted by chlorite along margins and fractures. (D) Euhedral olivine phenocryst completely pseudomorphed by calcite. (E) Interclast matrix assemblage of microlitic diopside (pale-yellow), and phlogopite (green – F), with an inhomogeneous distribution of apatite (green-yellow), calcite (bright yellow), and spinel (white).  Geological Unit Summaries - Kimb65a 29  The interclast matrix (icm) is the ultra-fine grained matrix that is host to country rock xenoliths, magmaclasts, and discrete crystals (figure 3.1). The interclast matrix contains subhedral to euhedral microlites of phlogopite and diopside with minor interstitial serpentine, with primary kimberlite groundmass spinel and apatite distributed in an irregular and inhomogeneous manner (figure 3.2E & 3.2F). Rocks in Kimb65a are typically matrix supported, but in some cases where microxenolith (<1 mm) dilution is particularly high, the rock may show a clast-supported texture. Primary kimberlite groundmass (gm) is observed only within the cored, uncored, and incipient magmaclasts (figure 3.1). The groundmass mineral assemblage, in order of decreasing modal abundance, includes: serpentine, calcite, phlogopite, apatite, spinel, perovskite, pyrite and titanite. The variability in modal abundance of groundmass minerals between magmaclasts is dependent on the degree of preservation of the magmaclasts. As magmaclasts become increasingly zoned (figure 3.1D & 3.2A), perovskite is partially replaced by Mn-Ilmenite, and clusters of subhedral to euhedral pyrite are common (figure 3.3). These features result in variability in the observed modal abundances of perovskite and spinel under petrographic microscope. Spinel is affected because the presence of fine-grained euhedral pyrite cannot be visually distinguished from spinel easily. Each groundmass mineral in Kimb65a is described below in terms of size, shape, distribution, textural relationships and modal abundance in the magmaclast groundmass: Serpentine (gm) accounts for 5 – 8 % modal abundance, forming anhedral segregations interstitial to the other groundmass minerals (figure 3.3 & 3.4). Serpentine is commonly altered to chlorite. Calcite (cal) accounts for 4 – 7 % modal abundance, forming anhedral segregations interstitial to the other groundmass minerals (figure 3.3 & 3.4) similar to serpentine.  Geological Unit Summaries - Kimb65a 30          Figure 3.3 - (A) Magmaclast hosted primary groundmass minerals observed in well preserved magmaclasts (figure 3.1A & 3.1F) from Kimb65a. Phlogopite, apatite, spinel, and perovskite are hosted within an interstitial groundmass mesostasis of chlorite, after serpentine, and calcite. Calcite pseudomorphs after olivine macrocrysts and euhedral microcrysts with minor serpentinization (B) Element distribution map using P (blue) to illustrate apatite, Cr (pink-purple) to illustrate magnesiochromite-spinel, Ti (orange) to illustrate perovskite. (C) Element distribution map using K (green) to illustrate phlogopite. Figure 3.4 - (A) Magmaclast hosted primary groundmass minerals observed in poorly preserved magmaclasts (figure 3.1B & 3.1E) from Kimb65a. In contrast to the well preserved magmaclasts (i) calcite pseudomorphs after olivine are replaced by serpentine and chlorite, (ii) former perovskite grains nearly absent, small residual partial Mn-ilmenite rims remain, and (iii) groundmass calcite replaced by chlorite and serpentine, small residual interstitial calcites remain (B) Element distribution map using P (blue) to illustrate apatite, Cr (pink-purple) to illustrate magnesiochromite-spinel, Ti (orange) to illustrate perovskite. (C) Element distribution map using Ca (yellow) to illustrate both apatite and interstitial calcite, and K (green) to illustrate phlogopite. Geological Unit Summaries - Kimb65a 31   Figure 3.5 – (A & B) BSE image of subhedral groundmass magnesiochromite spinels, with atoll texture showing a very thin atoll of ulvöspinel (white), separated by a mixture of calcite and serpentine. (C) Euhedral groundmass pyrite. (D) Anhedral aggregate of pyrite (white) hosting euhedral hexagonal crystals of apatite as inclusions (light-grey). (E) Groundmass perovskite replaced by Mn-ilmenite preserving the original euhedral crystal shape of the perovskite; perovskite appears partially resorbed with calcite filling the space between the perovskite and the Mn-ilmenite rim. (F) Anhedral interstitial titanite, partially hosting inclusions of olivine and phlogopite.   Geological Unit Summaries - Kimb65a 32  Phlogopite (gm) accounts for 12 – 15 % modal abundance, forming subhedral-euhedral lath-like crystals ranging in size from 5 – 100 µm homogeneously distributed throughout the groundmass (figure 3.3 & 3.4). Apatite (gm) accounts for 5 – 10 % modal abundance, forming subhedral-anhedral crystals ranging in size from 5 – 8 µm (Figure 3.3 & 3.4). Apatite distribution is concentrated within the interstitial mesostasis of serpentine and calcite.  Spinel (gm) accounts for 1 – 3 % modal abundance, forming euhedral crystals ranging in size from 5 – 20 µm (figure 3.3 & 3.4). These spinels have an atoll texture with a subhedral Mg-rich core of magnesiochromite and a thin (1 – 2 µm) outer-rim of ulvöspinel. The ulvöspinel is separated from the spinel core by calcite and serpentine (figure 3.5A & 3.5B).  Pyrite (gm) accounts for 1 – 2 % modal abundance, forming subhedral-euhedral crystals ranging in size from 15 – 75 µm. In most cases, pyrite is distributed in the groundmass as single crystals (figure 3.5C), although in some cases, it forms aggregates of subhedral crystals, partially trapping inclusions of apatite (figure 3.5D). Pyrite abundance shows a strong correlation with magmaclast size; larger magmaclasts contain abundant pyrite while smaller magmaclasts contain little to no pyrite. Perovskite (gm) accounts for 1 – 2 % modal abundance, forming subhedral crystals ranging in size from 10 – 25 µm. Perovskites appear to be pseudomorphed along the margins of the crystal by small euhedral crystals of Mn-ilmenite. The Mn-ilmenite maintains the original euhedral crystal shape of the perovskite (3.5D). While the original and partially destroyed perovskite crystal may still be present in the core, more commonly the Mn-Ilmenite rim is all that remains. Titanite (gm) accounts for <1 % modal abundance, forming interstitial anhedral 20 – 30 µm crystals partially enclosing olivine and phlogopite (figure 3.5E). Geological Unit Summaries - Kimb65b 33  3.1.2 Kimb65b The following description is a representative summary of a rock type with a consistent set of kimberlite features, repeated over large continuous intersections within the vertical drill holes R65-34 reaching 200 m depth and R6-01 reaching 200 m depth (figure 1.8). Visual examination of a further 4,000 m of drill core from 23 additional vertical and sub-vertical drill holes confirms that rocks consistent with the characteristics of Kimb65b form a vertical-conical body occupying approximately 10 % of the pipe by volume (figure 1.8).  Kimb65b can be described as a xenolith poor olivine-rich magmaclast-poor coherent to lesser transitional volcaniclastic kimberlite. The rocks display a coherent macrocrystic texture with subround country rock xenoliths, anhedral olivine macrocrysts, and euhedral olivine microcrysts (phenocrysts) in a fine-grained primary kimberlite groundmass (figure 3.6A). The coherent texture varies over meter-scale intervals in drill core to a transitional coherent to volcaniclastic texture with patches of diopside-rich interclast matrix between large interconnected incipient magmaclasts (figure 3.6B). Kimb65b shows a range of CR xenolith dilution from 15 – 30 %. There is a general correlation between higher localized xenolith abundances (20 – 30 %) with transitional textures (figure 3.6B), and lower localized xenolith abundances (15 – 20 %) with more coherent textures (figure 3.6A).  Country rock xenoliths (CRx) range in size from 0.5 mm to 4 cm xenocrysts and clasts, up to a maximum of 12 cm. The clasts are subround to round, and composed of a mixture of gneissic and granitoid protoliths with both primary and reaction mineral assemblages. Larger clasts (> 2 cm) are generally more zonally reacted, displaying very fresh cores of granite and gneiss, and 5 mm – 1 cm reaction rims with pale-green discolouration (figure 3.6C). Smaller (< 2 cm) xenoliths are very strongly uniformly reacted, with reaction mineral assemblages of calcite, serpentine, and pectolite replacing the cores, with successive outer rims of diopside + aegirine + serpentine, and phlogopite + apatite (figure 3.6 & 3.7). Geological Unit Summaries - Kimb65b 34   Figure 3.6 – (A) polished slab (PS) with coherent macrocrystic texture, abundant anhedral olivine macrocrysts and strongly reacted country rock xenoliths hosted in a vfg groundmass. The degree of olivine serpentinization is strongly correlated to proximity to strongly reacted CR xenoliths. (B) Coherent to lesser volcaniclastic inequigranular texture, fully serpentinized olivine macrocrysts, microcrysts and CR xenoliths in a vfg groundmass; groundmass is mottled with patches of microlitic diopside-rich interclast matrix. (C) Larger CR xenoliths display mild to moderate zonal reaction with kimberlite, preserving fresh cores. Reduced degree of reaction of the xenolith is correlated to low degrees of serpentinization of proximal olivine. (D) Coherent to lesser volcaniclastic texture highlighting the development of incipient magmaclasts, with irregular patches of interclast matrix. (E) plane-polarized light thin-section (PPL) of coherent kimberlite showing phlogopite phenocryst-rich groundmass surrounding CR xenoliths and olivine macrocrysts. Note that CR xenoliths display thick brown phlogopite-rich rims. (F) PPL of coherent to lesser volcaniclastic kimberlite showing phlogopite phenocryst-rich groundmass surrounding serpentinized olivine macrocrysts and CR xenoliths; note that CR xenoliths contain aegirine-rich inner rims. Geological Unit Summaries - Kimb65b 35   Figure 3.7 – (A) BSE image of CR xenoliths commonly observed in Kimb65b with coherent textures; xenolith cores are replaced by calcite, pectolite, and serpentine, with consecutive outer rims of serpentine and diopside, followed by phlogopite and apatite. (B) 5x zoom of xenolith rims from (A). (C) BSE image of CR xenoliths commonly observed in Kimb65b with coherent to lesser volcaniclastic texture; the xenolith cores are replaced by calcite, pectolite and serpentine, followed by a rim of serpentine and fine-grained diopside zoned to coarse-grained aegirine, and an outermost rim of phlogopite and apatite. (D) 3x zoom of xenolith rims from (C) showing texture of consecutive reaction rims. (E) PPL and (F) XPL image of xenoliths in coherent to lesser volcaniclastic kimberlite showing calcite and serpentine-rich cores, with a large inner rim of aegirine; aegirine crystals appear to be zonations on diopside crystals which decrease in grain size rapidly into the diopside and chlorite-rich outer rim.   Geological Unit Summaries - Kimb65b 36  The abundance of aegirine varies with kimberlite texture; xenoliths in rocks with a more coherent texture contain little or no aegirine zonation on the outer diopside rims, and more phlogopite along the outermost rims of the xenoliths (figure 3.6A, 3.6E, 3.7A & 3.7B). CR xenoliths in rocks with a more transitional volcaniclastic texture contain significantly more aegirine (giving these xenoliths a distinct green colour), and significantly less phlogopite along the outermost margin of the reaction rims (figure 3.6B, 3.6F, 3.7C, 3.7D, 3.7E & 3.7F). Olivine (ol) macrocrysts account for 20 – 25% modal abundance of the rock, forming anhedral sub-round crystals, ranging from 2 – 8 mm and commonly as large as 14 mm in size. Olivine phenocrysts account for 20 – 25 % modal abundance of the rock, forming subhedral to euhedral crystals, ranging in size from 0.25 – 1 mm. Olivine macrocrysts are well preserved, and may be partially or fully pseudomorphed by serpentine, particularly in the 1 mm – 2 cm vicinity of CR xenoliths (figure 3.6A). Olivine replacement by serpentine appears to progress from the margins of the crystal towards the inner portions of the crystal, and the serpentine contains abundant microscopic (1 – 2 µm) inclusions of millerite (NiS). This serpentine appears to cross-cut an earlier generation of more vein-like serpentine that seems to have affected even the freshest olivine macrocrysts (figure 3.6A and figure 3.6B). Olivine phenocrysts are not preserved, and have been fully pseudomorphed by serpentine. Magmaclasts (mc) as discrete clasts of solidified kimberlite magma (as observed in Kimb65a and Kimb65d) are absent in this unit; interconnected patches of coherent groundmass, separated by irregular patches of diopside-rich interclast matrix, define an incipient transitional magmaclastic texture over meter-scale intervals within the largely coherent kimberlite (figure 3.6B, 3.6D & 3.6F).  The primary kimberlite groundmass (gm) mineral assemblage, in order of decreasing modal abundance, includes: phlogopite, serpentine, calcite, perovskite, spinel, and apatite (figure 3.8). Where transitional textures are more dominant, the groundmass contains small isolated patches of diopside,Geological Unit Summaries - Kimb65b 37           Figure 3.8 - (A) Coherent Kimb65b groundmass with serpentinized olivine phenocrysts, phlogopite, interstitial calcite, and serpentine, plus apatite, spinel, perovskite, and monticellite. (B) Modal abundances of spinel (purple and green for magnesiochromite and ulvöspinel), perovskite (orange) and apatite (blue). (C) Modal abundance of phlogopite (green). Figure 3.9 - (A) Coherent to transitional volcaniclastic Kimb65b groundmass with serpentinized olivine phenocrysts, phlogopite, interstitial calcite, and serpentine, plus apatite, spinel, perovskite, and monticellite; transitional texture indicated by patches of microlitic diopside (B) Modal abundances of spinel (purple for magnesiochromite and ulvöspinel), perovskite (orange) and apatite (blue). (C) Modal abundance of phlogopite (green). Geological Unit Summaries - Kimb65b 38  and increased abundance of apatite (figure 3.9). Each mineral is described below in terms of size, shape, distribution, textural relationships and modal abundance in the groundmass. Phlogopite (gm) accounts for 15 – 20 % modal abundance, forming euhedral crystals ranging in size from 15 – 100 µm (figure 3.8A, 3.8C, 3.9A & 3.9C). Phlogopite is homogeneously distributed throughout the groundmass. Slightly larger phlogopite crystals are also observed forming rims around reacted CR xenoliths (figure 3.7B & 3.7D).  Serpentine (gm) accounts for 10 – 12 % modal abundance, forming interstitial anhedral segregation pools ranging from 20 – 50 µm across. Serpentine segregation pools contain inclusions of phlogopite, apatite, spinel and perovskite (figure 3.8A & 3.9A). Calcite (gm) accounts for 5 – 10 % modal abundance, forming interstitial anhedral segregation pools ranging from 50 – 150 µm across. These segregation pools often contain inclusions of phlogopite, apatite, spinel, and perovskite (figure 3.8A & 3.9A). Spinel (gm) accounts for 4 – 7 % modal abundance, forming zoned, subhedral crystals ranging in size from 10 – 30 µm. The cores of the spinels are rich in Mg and Cr (chromite) and the rims are rich in Fe and Ti (ulvöspinel) (figure 3.10A). The boundary between the two spinel compositions does not conform to the growth zones of the spinel. Smaller subhedral ulvöspinel crystals (5 – 10 µm) are commonly observed adjacent to or in contact with ulvöspinel mantles on chromites (figure 3.8A & 3.9A). Spinels are the only groundmass mineral that is observed as inclusions within olivine phenocrysts. Such inclusions are typically of a purely magnesiochromite composition without ulvöspinel mantles, in stark contrast to the spinels in the adjacent groundmass (figure 3.10C). Geological Unit Summaries - Kimb65b 39   Figure 3.10 – (A) Spinel with chromite core composition, and an ulvöspinel rim composition. The ulvöspinel rim does not conform to the growth faces of the chromite. Subhedral smaller ulvöspinel crystals occur adjacent to, and appear to fragment off of chromites. (B) Very strongly resorbed spinel; resorbtion affects chromite cores but does not affect ulvöspinel rims. (C) subhedral serpentinized olivine phenocryst, with inclusion of euhedral chromite; note that chromite in the adjacent groundmass has a mantle (in this case parallel to the chromite growth faces) of ulvöspinel, while the chromite inclusion does not. (D) Euhedral perovskite with multiple compositional zonations. Core compositions are homogeneous over approx. 15 µm, with a darker 5 µm rim; small 1 µm irregular zonations occur in this outer rim. Subhedral ulvöspinel commonly occurs as inclusions within these outer compositional rims. (E) Interclast matrix mineral assemblage is dominated by microlites of phlogopite > diopside. Euhedral and well preserved crystals of apatite, spinel and perovskite are irregularly distributed throughout the interclast matrix in low 1 – 3 % modal abundances. (F) Element distribution map to aid in visualizing the interclast matrix modal mineral assemblages: phlogopite (K, green), diopside (Ca, yellow), apatite (Ca, P, yellow + blue), spinel (white), serpentine (grey). Geological Unit Summaries - Kimb65b 40  Perovskite (gm) accounts for 2 – 4 % modal abundance, forming subhedral to euhedral crystals ranging from 20 – 40 µm in size. Perovskite cores are relatively homogeneous in composition, but ultra- thin 1 µm zonation lines, discordant with the crystal growth zones, are observed within the outer margins of the crystal; subhedral ulvöspinel is observed as partial inclusions within the margins of the perovskite crystal (figure 3.10D).  Apatite (gm) accounts for a highly variable 1 – 12 % modal abundance, forming subhedral to euhedral acicular crystals ranging in size from 5 – 100 µm (figure 3.8A & 3.8B). The modal abundance of apatite is closely correlated with changes in texture from coherent to transitional volcaniclastic; apatites become much more abundant as the kimberlite texture becomes more transitional (figure 3.9A & 3.9B). The range in size is mostly attributable to the acicular habit.  The interclast matrix (icm) forms on a millimeter to centimeter scale (figure 3.6B, 3.6D & 3.6F), consisting of patches containing abundant microlitic phlogopite, lesser diopside, interstitial serpentine, and variable abundances of serpentinized olivine phenocrysts, apatite, spinel and perovskite with an inhomogeneous distribution (figure 3.10E & 3.10F).         Geological Unit Summaries - Kimb65c 41  3.1.3 Kimb65c The following description is a representative summary of a rock type with a consistent set of kimberlite features, repeated in 1 cm to 10 m intersections with sharp contacts cross-cutting kimberlite and non-kimberlite geological units within two vertical drill-holes (R65-31 & R65-33) reaching 191 m and 200 m depth (figure 1.8). Visual examination of a further 4,000 m of drill core from 23 additional vertical and sub-vertical drill holes confirms that rocks consistent with the characteristics of Kimb65c form small sheet-like bodies occurring sporadically throughout the pipe and adjacent country-rock, occupying approximately 2% of the pipe by volume; they are not shown on the geological model (figure 1.8) due to their small size and widespread distribution. These HK dykes cross-cut kimberlite and non-kimberlite geological units (figure 3.11E), and appear to be most abundant near contacts between geological units (figure 3.12). It should be noted that visual examination of Kimb65c intersections within the other 36 drill-holes, indicates some variability in olivine macrocryst size and replacement, an indication that perhaps rocks classified as Kimb65c include more than one phase of kimberlite. However, a detailed characterization of the total number of kimberlite phases represented by Kimb65c requires significant representative sampling across a large number of drill-holes and is beyond the scope of this study.  Kimb65c can be described as a xenolith-poor olivine-rich macrocrystic coherent kimberlite. Rocks display a macrocrystic coherent texture with calcite-pseudomorphed olivine macrocrysts and phenocrysts, and a very low abundance of strongly reacted country rock xenoliths, set in a primary kimberlite groundmass.  Geological Unit Summaries - Kimb65c 42   Figure 3.11 – (A) Coherent texture with calcite-pseudomorphed olivine macrocrysts and phenocrysts, variably overprinted by serpentine, and strongly reacted country rock xenoliths, in very fine-grained crystalline groundmass. (B) Plane polarized light (PPL) image showing calcite pseudomorphed olivine macrocrysts, partially replaced by serpentine, within a vfg groundmass. (C) Close-up of a calcite pseudomorphed olivine macrocryst showing serpentine replacement along the cleavage planes of the calcite. (D) PPL image showing large country rock xenolith in reaction with host kimberlite. (E) Typical occurrence of Kimb65c in drill core as small sheet-like bodies cross-cutting both kimberlite and non-kimberlite geological units, with sharp parallel contacts.      Geological Unit Summaries - Kimb65c 43   Figure 3.12 – Common occurrence of Kimb65c hypabyssal dyke near the contact between coherent to volcaniclastic Kimb65b and volcaniclastic Kimb65a. Contacts between Kimb65b and Kimb65a, and contacts between Kimb65b and Kimb65c are all sharp, and planar. Olivine (ol) macrocrysts account for 20 – 30 % modal abundance, ranging in size from 1 – 4 mm and commonly up to 8 mm in size, with anhedral crystal shapes. Olivine phenocrysts account for 15 – 20 % modal abundance, ranging in size from 0.5 – 1 mm in size, with subhedral crystal shapes. Olivine macrocrysts and phenocrysts are entirely pseudomorphed by calcite, and subsequently overprinted by serpentine along the margins of the crystal, and along the cleavage planes of the calcite (figure 3.11C). These macrocrysts and microcrysts are determined to have been olivine prior to replacement on the basis of their size, shape, abundance, and distribution.    The primary kimberlite groundmass (gm), in order of decreasing modal abundance, include:  serpentine, calcite, apatite, phlogopite, pyrite, perovskite, and spinel (figure 3.13). Each mineral is described below in terms of size, shape, distribution, textural relationships and modal abundance in the groundmass. Geological Unit Summaries - Kimb65c 44  Serpentine (gm) accounts for 10 – 12 % modal abundance, forming interstitial anhedral segregations (light-grey interstitial material) between the ultra-fine-grained groundmass calcite and phlogopite (figure 3.13C).   Calcite (gm) accounts for 7 – 10 % modal abundance, occurring as anhedral crystals 1 – 3 µm in size, hosted in interstitial serpentine (figure 3.13C).  Phlogopite (gm) accounts for 5 – 10 % modal abundance, forming homogeneously distributed euhedral crystals ranging in size from 10 – 25 µm (figure 3.13A & 3.13C). Pyrite (gm) accounts for 3 – 6 % modal abundance, forming anhedral globular to subhedral acicular crystals ranging from 5 – 40 µm in length. The distribution of pyrite is either (i) larger 30 – 30 µm single crystals (figure 3.13A) (ii) clusters of 5 – 10 µm crystals (figure 3.14E) (iii) anhedral 1-2 µm blebs within titanite grains (figure 3.14D) and (iv) clusters of 1 – 5 µm crystals mantled around anhedral Mn-Ilmenite grains (figure 3.14E). Perovskite (gm) accounts for 2 – 5 % modal abundance, forming anhedral crystals homogeneously distributed throughout the groundmass (figure 3.22 & 3.23D). In almost all cases, perovskite crystals appear to be replaced by mixtures of titanite, Mn-ilmenite and pyrite (figure 3.14D & 3.14E). The rare occurrences of preserved perovskite cores mantled by Mn-ilmenite (figure 3.14C) demonstrate that these mixtures of titanite, pyrite and Mn-ilmenite represent former perovskites. Spinel (gm) accounts for 1 – 2 % modal abundance, forming subhedral to euhedral crystals ranging in size from 1 – 11 µm. Somewhat unusually, the vast majority of spinels are hosted as inclusions within former olivine phenocrysts, with only a trace abundance of spinels hosted directly in the kimberlitic groundmass (figure 3.22 & 3.23B).  Geological Unit Summaries - Kimb65c 45      Figure 3.13 – (A) Coherent Kimb65a groundmass mineral assemblage containing Mn-Ilmenite (after perovskite), spinel, apatite, pyrite, phlogopite, and calcite. (B) Element distribution map using Ti (orange) to show Mn-ilmenite after perovskite, Cr (purple-pink) to show spinel, P (baby blue) to show apatite, and S (purple-blue) to show pyrite. (C) Element distribution map using Ca (yellow) to show calcite, K (green) to show phlogopite, and P (baby blue) to show apatite; note that P is mapped to visually separate apatite from calcite. Figure 3.14 – (A) Serpentinized olivine phenocryst containing an inclusion of magnesiochromite. (B) olivine phenocrysts are generally completely serpentinized, but in some cases (as with the olivine macrocrysts) calcite cores are not overprinted by serpentine. (C) Groundmass perovskite with a thin mantle of Mn-ilmenite; the vast majority of groundmass perovskite crystals are fully replaced (see D and E). (D) Anhedral titanite with blebs of pyrite after perovskite. (E) Anhedral to acicular pyrite grains mantled around Mn-ilmenite. (F) ultra fine-grained groundmass phlogopite with anhedral calcite and interstitial serpentine.  Geological Unit Summaries - Kimb65d  46  3.1.4 Kimb65d The following description is a representative summary of a rock type with a consistent set of kimberlite features, repeated over large continuous intersections within the vertical drill hole R6-01 reaching 200 m depth and R65-34 reaching 200 m depth (figure 1.8). Visual examination of a further 4,000 m of drill core from 23 additional vertical and sub-vertical drill holes confirms that rocks consistent with the characteristics of Kimb65d form a vertical-conical body occupying approximately 25 % of the pipe by volume (figure 1.8).  Kimb65d can be described as a xenolith-rich olivine-poor magmaclast-rich volcaniclastic kimberlite. Rocks display an inequigranular and volcaniclastic texture, with subround country rock xenoliths, discrete pseudomorphed olivine macrocrysts, and a magmaclastic texture defined by (i) cored, (ii) uncored, and (iii) rarely incipient magmaclasts, all hosted within an interclast matrix composed of microlitic phlogopite and diopside, and irregularly distributed primary kimberlite groundmass minerals (figure 3.15A, 3.15B & 3.15C).  Country rock xenoliths (CRx) range in size from 0.5 mm to 5 cm up to a maximum size of 12 cm with a range in modal abundance from 20 – 40 %. Clasts are subround and display a very distinctive macroscopic green colour with a thin (2 mm) light-beige outer rims (figure 3.15A). Coarser clasts (>2 cm) generally maintain their primary mineral assemblage within the core of the clast, developing a sequence of reaction rim mineral assemblages towards the clast margins; smaller (<2 cm) clasts do not show preserved primary minerals in the clast core. The xenolith reaction mineral assemblage is dominated by anhedral calcite, serpentine and pectolite throughout the core of the clast (figure 3.15D, 3.15E & 3.15F). Towards the margins of the clast are sequential rims of aegirine (with crystals oriented perpendicular to the clast margins) grading into much finer-grained diopside (mixed with lesser chlorite) and a very thin outermost rim of phlogopite with inclusions of anhedral apatite. The abundance of aegirine and the   Geological Unit Summaries - Kimb65d  47   Figure 3.15 – (A) Polished slab (PS) with volcaniclastic to transitional volcaniclastic texture, showing a round magmaclast partially cored on a serpentinized olivine phenocryst. Magmaclasts, zonally reacted country rock xenoliths and xenocrysts, and serpentinized olivine macrocrysts are hosted within an interstitial interclast matrix. (B) Plane polarized light image showing a thick-selvage magmaclast cored on a serpentinized olivine macrocryst. (C) Plane polarized light image showing a thin-selvage magmaclast cored on a serpentinized olivine macrocryst. (D) Cross-polarized light image showing the typical zonations of country rock xenolith reaction mineral assemblage, with anhedral calcite and serpentine in the core of the xenolith, followed by a rim of relatively coarse aegirine crystals, followed by a finer-grained rim of diopside. (E) Backscattered electron image showing less common xenolith reaction mineral assemblages that do not contain aegirine. Xenolith cores are composed of anhedral serpentine, calcite and diopside, followed by a rim of ultra-fine-grained serpentine and diopside, and a very thin outer rim of diopside, followed by an outermost rim of phlogopite and apatite. A higher magnification image of the outermost portions of the reaction rim are displayed in (F).   Geological Unit Summaries - Kimb65d  48  relatively thick outer rims of diopside give xenoliths in Kimb65d their distinctive dark-grey to pale-green colour with light-beige outer rims.  Olivine (ol) macrocrysts account for 15 – 20 % modal abundance, ranging in size from 2 – 8 mm with anhedral and round crystal shapes. Olivine phenocrysts account for 10 – 15 % modal abundance, ranging in size from 0.25 – 1 mm with subhedral crystal shapes. Olivine macrocrysts and phenocrysts are not preserved in this geological unit, and in all cases have been pseudomorphed by serpentine with abundant 1 – 2 µm inclusions of millerite (figure 3.16A). The margins of these former olivine grains, and along fractures, show alteration of serpentine to chlorite (figure 3.16A).  Magmaclasts are abundant in this unit, ranging in size from 3 mm – 1 cm with a typical modal abundance of 5 – 10 %. The magmaclasts have round to subround clast shapes, typically forming a thick-selvage, cored on an olivine macrocrysts or country rock xenolith (figure 3.15A, 3.15B & 3.15C). In some cases, these magmaclasts become more irregular or patchy, and show a mild degree of interconnectivity with surrounding magmaclasts, and are thus described as incipient magmaclasts (figure 3.15A – reddish brown area at the bottom right of the image). Primary kimberlite groundmass (gm) is preserved within magmaclasts, and is generally consistent between different magmaclasts at different depths in the unit. The groundmass mineral assemblage in order of decreasing abundance includes: phlogopite, serpentine, perovskite, spinel, and apatite (figure 3.16). Each mineral is described below in terms of size, shape, distribution, textural relationships and modal abundance in the groundmass.  Phlogopite (gm) accounts for 20 – 25 % modal abundance, forming euhedral crystals ranging in size from 15 – 50 µm with an inner and outer compositional zonation (figure 3.16, 3.17A, 3.17B & 3.17D).  Geological Unit Summaries - Kimb65d  49   Figure 3.16 – Magmaclast hosted primary kimberlite groundmass containing phlogopite, serpentine, spinel, perovskite and apatite. Olivine macrocrysts are fully serpentinized with small inclusions of millerite, and overprinted by chlorite along the margins of the crystal. (B) Element distribution map showing the abundance and distribution of groundmass magnesiochromite-ulvöspinel (purple – Cr), perovskite (Ti – orange) and apatite (P – blue). (C) Element distribution map showing distribution of phlogopite (K – green) and interstitial serpentine (dark grey interstitial material).    Geological Unit Summaries - Kimb65d  50   Serpentine (gm) accounts for 8 – 10 % modal abundance, forming interstitially to all other groundmass minerals, as well as slightly larger (20 – 25 µm) segregation pools (figure 3.16).  Perovskite (gm) accounts for 5 – 7 % modal abundance, forming subhedral to euhedral crystals ranging in size from 7 – 30 µm (figure 3.16B). Perovskites do not show any compositional zonation (figure 3.17B).  Spinel (gm) accounts for 3 – 6 % modal abundance, forming zoned crystals with euhedral magnesiochromite cores and thin anhedral ulvöspinel rims, ranging in size from 10 – 45 µm (figure 3.16). While the ulvöspinel rims of these crystals are well preserved, the magnesiochromite cores are often partially resorbed, and in some cases almost fully resorbed, forming an atoll spinel texture (figure 3.17A).   Apatite (gm) accounts for <<1 % modal abundance, forming anhedral to occasionally subhedral crystals ranging from 2 – 10 µm in size (figure 3.16). These apatites commonly contain <1 µm impurities (figure 3.17C).  The interclast matrix (gm) is composed primarily of euhedral zoned phlogopite, with lesser subhedral microlitic diopside, and variable abundances of reacted CR xenolith and xenocrysts fragments, as well as primary perovskite and spinel (figure 3.31).   Geological Unit Summaries - Kimb65d  51   Figure 3.17 – (A) Groundmass spinel with magnesiochromite core, with a very thin mantle of ulvöspinel. The magnesiochromite cores appear to show a very thin compositional zonation within the outer 1 µm of the crystal cores. The resorbtion of these magnesiochromite cores and the development of atoll textures are common. Also note compositionally zoned phlogopite in the adjacent groundmass. (B) Euhedral perovskite that does not show compositional zonation. Note again the compositionally zoned phlogopite in the adjacent groundmass. (C) Very small subhedral apatite crystals riddled with <1 µm inclusions. (D) Compositionally zoned groundmass phlogopite. (E) Interclast matrix dominated by phlogopite with lesser microlitic diopside, with heterogeneous abundances of groundmass minerals including spinel, perovskite and apatite. (F) Element distribution map showing the distribution and relative abundances of interclast matrix phlogopite (K – green), diopside (Ca – yellow), spinel and perovskite (white).   Geological Unit Summaries - CR, CCR, CRB  52  3.1.5 CR, CCR, CRB  Country-rock (CR) is a geological unit, not included in the geological model, referring to kimberlite host-rocks from the Archean basement gneissic-granitoids of the eastern Superior Craton. These rocks are variable between coarse-grained granitoids composed of primarily albite, alkali-feldspar and quartz with minor biotite and garnet, to gneissic foliated rocks composed of albite, quartz, and biotite (figure 3.1). In drill-core, country rock xenoliths larger than 1 m, within kimberlite, are assigned to the CR geological unit; otherwise the only occurrence of this unit is approximately 10-50 m from the kimberlite pipe walls; separated from the pipe by zones of fractured and brecciated country rock. Fractured country-rock, termed ‘cracked country rock’ (CCR) in the geological model, refers to a geological unit that is consistent in composition with CR, but shows significant fracturing with kimberlite material infilling fractures. Fractures range in thickness from 1 mm – 5 cm, occurring approximately every 15 – 20 cm, and are infilled with finely-pulverized angular country-rock fragments (0.25 – 2 mm); the matrix between these rock fragments is frequently composed of serpentine, less common carbonate, and rare serpentinized olivine crystals. This rock type is encountered on a meter-scale within a 5 – 40m zone around the kimberlite pipe, and on a centimeter-scale within a 10-30 cm zone surrounding kimberlite dykes. In both cases, the fractures containing primary kimberlite components are interpreted to result from kimberlite emplacement-related processes. Country-rock breccia (CRB), is a geological unit composed of clast-supported sub-angular country-rock clasts, in a matrix of angular pulverized country-rock fragments (0.25-2 mm), abundant serpentine, frequent serpentinized olivine (<2 mm) and occasional carbonate. This geological unit is generally encountered within a 1 – 20 m zone around the kimberlite pipe, and is typically in contact with kimberlitic rocks in the diatreme of the pipe. The < 95 % modal abundance of country rock material in these rocks differentiates them from highly CR xenolith diluted volcaniclastic kimberlite.  Geological Unit Summaries - CR, CCR, CRB  53   Figure 3.18 – (A) Plane polarized light image showing typical mineralogy of ‘gneissic’ country rock xenoliths, containing quartz and feldspar with a foliation defined by the parallel alignment of abundant biotite crystals. (B) Plane polarized light image showing typical mineralogy of ‘granitic’ country rock xenoliths, containing inequigranular feldspar, quartz, and biotite. (C) Drill core photograph showing typical fractured country rock (CCR) adjacent to the pipe, with 1 mm – 5 cm fractures infilled with pulverized country rock fragments, in a finer matrix of serpentine, less common carbonate, and rare serpentinized olivine. (D) Drill core photograph showing typical brecciated and mobilized country rock (CRB) adjacent to the pipe. Angular country rock clasts ranging in size from 0.5 mm fragments up to 10 cm clasts, account for 95 % of the modal abundance of the rock. Clasts appear subangular, and the matrix between the country rock includes serpentine, rare carbonate, and rare serpentinized olivine.  Results - Mineral Chemistry  54  3.2 Mineral Chemistry Mineral compositions are reported for groundmass mineral assemblages, interclast matrix mineral assemblages and xenolith reaction mineral assemblages in all kimberlite geological units of Renard 65. Compositions are subdivided by kimberlite rock texture, geological unit, and textural position. Abbreviations are: country rock xenolith (CRx), groundmass (GM), interclast matrix (ICM), magmaclast (MC), macrocryst (Macro), pseudomorph after olivine (PO), and not detected (nd). Compositions are reported as averages of minerals from the same textural position in the same geological units, with the number of analyses included. The complete non-averaged analyses and minimum detection limits are presented in Appendix C – Renard 65 Mineral Chemistry. 3.2.1 Olivine Olivine at one point constituted between 15 - 50 % modal abundance of any of the kimberlite units of Renard 65. However, in most of the preserved rock types the anhedral to euhedral olivines are almost always pseudomorphed either by calcite, or serpentine. The only preserved olivine encountered in this study is from HK textural varieties of Kimb65b. Olivine compositional data is therefore only reported for olivine macrocrysts hosted within Kimb65b (Table 3.2). These olivine macrocrysts have a Mole % Forsterite of Fo92.       Results - Mineral Chemistry  55  Table 3.2 – Compositions of olivine macrocrysts in coherent rocks from Kimb65b  3.2.2 Carbonate  Carbonates are observed in three textural positions (i) primary groundmass and magmaclast mineral assemblages, (ii) pseudomorphing after olivine macrocrysts and phenocrysts in Kimb65a and Kimb65c, and (iii) replacing primary xenolith mineralogy in reaction mineral assemblages in Kimb65b and Kimb65d. All of the carbonates encountered in Renard 65, regardless of geological unit or textural position, are calcites, with variable enrichments of Mg, Sr and Fe (Table 3.3).  Results - Mineral Chemistry  56  In Kimb65a, calcite pseudomorphing after olivine is essentially pure in composition, with minor Mn (0.33 wt% MnO), while groundmass carbonate is Sr-rich (0.41 wt% SrO) with minor Mg (0.36 wt% MgO), Mn (0.26 wt% MnO) and Fe (0.16 wt% FeO).  In the HK textural end-member of Kimb65b, groundmass calcite is very Sr-rich (1.32 wt% SrO) and Mg-rich (1.11 wt% MgO) with minor Fe (0.37 wt% FeO), while calcite replacing CR xenoliths is slightly less Sr-rich (0.58 wt% SrO) and Mg-rich (0.45 wt% MgO) with slightly more Fe (0.49 wt% FeO). In the HKt textural end-member of Kimb65b, similar but lower enrichment trends are observed for Sr (0.63 wt% SrO), Mg (0.37 wt% SrO) and Fe (0.18 wt% FeO) in the groundmass, and Sr (0.16 wt% SrO) in the CR xenoliths. Calcites from both groundmass and country rock xenoliths show Sr, Mn and Fe enrichment; enrichment is higher in groundmass calcites compared to country rock xenolith calcite, and enrichment is higher in HK textural end-members than in HKt textural end-members.            Table 3.3 – Averaged calcite compositions  Results - Mineral Chemistry  57  In Kimb65c, calcite pseudomorphing after olivine is essentially pure with minor Mn (0.13 wt% MnO) and Mg (0.11 wt% MgO), while calcite hosted in the groundmass is enriched in Sr (0.39 wt% SrO) with minor Mg (0.32 wt% MgO) and trace Fe (0.13 wt%). It is interesting to note the similar compositions of groundmass calcite and calcite pseudomorphing after olivine in both Kimb65a and Kimb65c. In Kimb65d, calcite is observed only as a replacement mineral in CR xenoliths, and is found to be nearly pure calcite with minor enrichment in Sr (0.16 wt% SrO).  3.2.3 Phlogopite Phlogopite is the most abundant mineral in both the kimberlite groundmass, and the interclast matrix, in all of the kimberlite units in Renard 65. In Kimb65b and Kimb65d, phlogopite crystals also form the outermost portion of reaction rims around country rock xenoliths. In most cases phlogopite crystals are not compositionally zoned and the analyses were obtained from the cores of the crystals; however, in Kimb65d, phlogopites show compositional zoning and both core and rim compositions are presented (Table 3.4).  In Kimb65a, phlogopite in the groundmass is enriched in Ti (1.10 and wt% TiO2) and Ba (0.98 and wt% BaO). Phlogopite hosted in the interclast matrix shows a similar but higher enrichment in Ti (1.62 wt% TiO2) and Ba (1.92 wt% BaO).  In Kimb65b, phlogopite in the groundmass is enriched in Ti (1.71 wt% TiO2) and Ba (1.15 wt% BaO), while phlogopite in the interclast matrix shows only enrichment in Ti (1.71 wt% TiO2). Phlogopite crystallizing along the outer-rims of CR xenoliths is enriched in Ti (1.93 wt% TiO2) and Ba (0.93 wt% BaO), similar in composition to phlogopite in the groundmass. In Kimb65c, phlogopite is observed only within the groundmass, and shows enrichment in Ti (0.57 wt% TiO2) and Ba (1.25 wt% BaO).  Results - Mineral Chemistry  58  Table 3.4 – Averaged phlogopite compositions    In Kimb65d, phlogopite crystals observed within the groundmass and interclast matrix show compositional zoning with core and rim compositions (figure 3.30A). Within the groundmass, phlogopite shows enrichment of Ti that decreases from core to rim (0.82 to 0.31 wt% TiO2) and enrichment of Ba that decreases from core to rim (0.53 to 0.26 wt% BaO). Additionally, these phlogopites show overall depletion of Al, with increasing depletion from core to rim (9.12 to 4.56 wt% Al2O3) and overall enrichment of Fe, Results - Mineral Chemistry  59  with increasing enrichment from core to rim (8.63 to 11.99 wt% FeO). Nearly identical compositions and zonation trends are observed for phlogopites hosted within the interclast matrix; they show enrichment of Ti that decreases from core to rim (0.82 to 0.18 wt% TiO2), enrichment of Ba that decreases from core to rim (0.52 to 0.13 wt% BaO), overall depletion in Al with increasing depletion from core to rim (9.42 to 3.54 wt% TiO2) and overall enrichment of Fe with increasing enrichment from core to rim (7.96 to 13.05 wt% Al2O3).  3.2.4 Spinel Spinel is one of the two most common oxide minerals in kimberlite groundmass, and magmaclasts in Renard 65. It is also likely one of the two first minerals (along with olivine) to crystallize from a kimberlite melt, as it is often found as inclusions in olivine phenocrysts. Averaged core compositions of spinels from groundmass and magmaclasts in each geological unit are presented in Table 3.5; rim compositions were not collected due to small crystal sizes (10 – 20 µm). Each geological unit includes more than 30 analyses taken at no less than 3 separate depths within the geological units; therefore, the compositional fields defined by these groups are spatially meaningful. The majority of kimberlite spinels crystallize in solid-solution within an 8 component system: MgCr2O4 (magnesiochromite) – FeCr2O4 (chromite) – MgAl2O4 (spinel) – FeAl2O4 (hercynite) – Mg2TiO4 (magnesian ulvöspinel) – Fe2TiO4 (ulvöspinel) – MgFe2O4 (magnesioferrite) – Fe3O4 (magnetite). Kimberlite spinel compositions can generally be graphically illustrated using a bivariate plot for the four most common end members: magnesiochromite, chromite, spinel, hercynite. Though in some cases, significant substitution of Ti or Fe3+ in the M2 site occurs, which may skew the representation of these compositions in a bivariate plot. To accommodate, two spinel prisms can be constructed (figure 3.19) modelled after those of Haggerty (1976).  Results - Mineral Chemistry  60  Table 3.5 – Averaged spinel compositions  Results - Mineral Chemistry  61   Figure 3.19 – Graphical presentations for kimberlitic spinels crystallized in either reduced or oxidized environments within the 8-component spinel prism after Haggerty (1976). Bi-variate plots illustrated for Renard 65 spinels are 2-dimensional projections into each prism from directions 1 and 2.  Results - Mineral Chemistry  62  The first prism assumes a more reduced melt, where substitution of Ti in the M2 site may be more dominant. The apex of the reduced prism is represented by the magnesian ulvöspinel and ulvöspinel end-members. The second prism assumes a more oxidized melt, where a significant amount of the Fe will be in the Fe3+ oxidation state. In this environment Fe3+ may substitution in the M2 site may be dominant. The two prisms, although designed to represent varied redox conditions, are not mutually exclusive. Spinels may have both Ti and Fe3+ so spinels should be rendered in both plots to ensure meaningful comparison of compositional trends.  Renard65 groundmass spinel compositions are illustrated in figure 3.20 in bi-variate plots projected from the 2-dimensional (1) Ti/(Ti+Cr+Al) vs. FeT/(FeT+Mg), and (2) Cr/(Cr+Al) vs. FeT/(FeT+Mg) surfaces of the reduced spinel prism, and the 2-dimensional (1) Fe3+/(Fe3++Cr+Al) vs. Cr/(Cr+Al), and (2) Fe2+/(Fe2++Mg) vs. Cr/(Cr+Al) surfaces of the oxidized spinel prism (figure 3.32). Projections into Ti/(Ti+Cr+Al) space in the reduced prism (figure 3.20A), and into Fe3+/(Fe3++Cr+Al) space in the oxidized spinel prism (figure 3.20B), both demonstrate that Renard 65 spinels are consistently and equivalently Ti-poor (approximately 2 – 4 wt% TiO2) and crystallized under reduced conditions with low Fe3+ content (approximately 4 – 6 wt% Fe2O3). As a result of these low Fe3+ and Ti mole % end-members, it is reasonable to identify compositional trends within illustrations of the bivariate Cr/(Cr+Al) – FeT/(FeT+Mg) plots either in the reduced or oxidized spinel prism. Considering that most spinels are believed to crystallize from kimberlite melts under relatively reduced conditions (Fedortchouk & Canil, 2004; Mitchell, 1995; Bellis & Canil, 2006), compositional trends will be identified in the reduced spinel prism with all Fe assumed to be present as Fe2+ (figure 3.34).  Results - Mineral Chemistry  63   Figure 3.20 – Renard 65 groundmass spinel compositions illustrated using bi-variate 2-d projections into the reduced spinel prism of figure 3.32. Results - Mineral Chemistry  64   Figure 3.21 – Spinel core compositions for each geological unit, illustrated using the Cr/Cr+Al vs. FeT/FeT+Mg bivariate plot. Kimb65b* refers to spinel compositions from the HKt textural variety of Kimb65b, while Kimb65b refers to spinel compositions from the HK textural variety. All of the spinels in the Renard 65 kimberlites are dominated by the magnesiochromite and chromite solid solution end members. Chromites from Kimb65a, regardless of whether they were sampled from a possible autolith (figure 3.1A & 3.1F, and labelled as LC for large clast in Table 3.5), from a well preserved magmaclast (figure 3.1B & 3.1E), or a poorly preserved magmaclast (figure 3.1B), show complete compositional overlap at an Fe # of 0.39 – 0.45 and a Cr # of 0.75 – 0.84.  Interestingly, the spinel compositions for Kimb65c show nearly perfect compositional overlap with Kimb65a, with a few high and low Cr # outliers. Spinels from the HK textural variety of Kimb65b have much higher Fe #’s (0.59 – 0.67) at Results - Mineral Chemistry  65  similar Cr #’s compared to the other units and show a very tightly constrained spatial distribution. Spinels from the HKt textural variety of Kimb65b (illustrated as Kimb65b* in figure 3.21) show an incredible range in compositions overlapping those of Kimb65b and Kimb65d from Fe # 0.41 – 0.65 at similar Cr #’s. Spinels from the KPKt textural variety of Kimb65d have Fe #’s from 0.40 – 0.49, with the majority of those spinels having Fe #’s below 0.45. They also have similar Cr #’s compared to the other geological units.  To summarize, spinel compositions from each of the geological units define compositional trends that overlap, partially overlap, or do not overlap with other geological units. In the case of Kimb65a, no differences in composition are observed for spinels hosted within magmaclasts of varying degrees of preservation, or within possible autoliths. Spinels hosted within the groundmass of Kimb65c, show complete compositional overlap with spinels from Kimb65a (correlating with the similar groundmass modal mineralogy between these two geological units). Compositional trends for groundmass hosted spinels in HK varieties of Kimb65b are far more Fe-rich than spinels in magmaclast hosted KPKt varieties of Kimb65d. However, spinels hosted within groundmass of HKt varieties of Kimb65b, show a large range of Fe #’s overlapping with trends defined by both Kimb65b and Kimb65d. 3.2.5 Perovskite  Perovskite is the second of the two most common groundmass oxides, crystallizing after spinel and olivine (Mitchell, 1973). In Renard 65 crystals have homogenous core compositions greater than 20 µm and in some cases irregularly zoned rims. All perovskite analyses reported are core compositions for groundmass or magmaclast hosted crystals in Kimb65a, Kimb65b and Kimb65d (Table 3.6). Crystals in Kimb65c are, to a greater extent than in Kimb65a, replaced by Mn-ilmenite (figure 3.5E) and intergrowths of titanite and pyrite (figure 3.4C). Perovskite analyses for Kimb65c are not reported due to insufficient preservation. Mn-ilmenite rims, and other replacement products of perovskite were not analyzed due to small (≤ 1 µm) crystal sizes. Results - Mineral Chemistry  66    The mineral chemistry of perovskites for Kimb65a, Kimb65b and Kimb65d (table 3.6) show similar enrichment, and very little compositional variation between the geological units. Substitutions of Ca2+ and Ti4+ are principally by coupled substitution or the creation of site vacancies by Fe3+, Nb5+ and REE3+. All perovskites show significant enrichment of Na (0.32-0.29 wt% Na2O), Fe3+ (2.41 – 2.84 wt% Fe2O3), Nb (1.20 – 1.27 wt% Nb2O5), Sr (0.23 – 0.27 wt% SrO) and REEs La (1.06 – 1.16 wt% La2O3) and Ce (2.54 – 3.22 wt% Ce2O3), and minor enrichment of Mg (0.10 – 0.11 MgO) and Al (0.52 – 0.57 wt% Al2O3). Low analytical totals are likely due to a limited REE analysis, as many kimberlitic perovskites contain 2 – 16 wt% total REE (Mitchell, 1986).  Table 3.6 – Averaged perovskite compositions Results - Mineral Chemistry  67  3.2.6 Apatite  Apatite occurs within the groundmass of all kimberlite units, and within the interclast matrix and as reaction minerals within CR xenoliths of Kimb65b and Kimb65d. Representative mineral compositions for apatites across textural positions in each geological unit are presented in Table 3.7. Apatite analyses from the rims of country rock xenoliths in Kimb65d were too small to produce reliable results, and their compositions are not reported. Apatites from the groundmass of Kimb65a are also too small for reliable results (wt% total ~ 89%) but their compositions are reported. Apatites show no significant compositional variation when compared across textural positions within a single geological unit, within a single textural variety. However, Sr enrichment increases in all textural positions (0.76 – 2.38 wt% SrO) from the HK endmember of Kimb65b to the HKt end-member. Similar Sr enrichment (1.20 – 1.98 wt% SrO) is observed within apatites in all textural positions of Kimb65d. There does not appear to be a correlation between textural progression and Sr-enrichment, as Kimb65a and Kimb65c show similarly minor Sr (0.70 and 0.90 wt% SrO).   Table 3.7 – Averaged apatite compositions Results - Mineral Chemistry  68  3.2.7 Pyroxenes and Pectolite Diopside, aegirine, and pectolite occur in variable quantities within country rock xenoliths as part of a reaction mineral assemblage in the geological units Kimb65b and Kimb65d. Diopside is also the second most abundant microlitic mineral in the interclast matrix of Kimb65a, Kimb65b and Kimb65d. Average compositions for these three pyroxenes in all available textural positions were collected primarily to provide confirmation of mineral assemblages, and comparison of diopside compositions between country rock xenoliths and the interclast matrix (Table 3.8).          Table 3.8 – Averaged compositions of diopside, aegirine and pectolite  Results - Mineral Chemistry  69  Aegirine is a country rock xenolith reaction mineral present in the reaction rims of xenoliths hosted within Kimb65b and Kimb65d. In Kimb65b, aegirine is relatively pure in composition, with minor enrichment in Ti (0.78 – 1.45 wt% TiO2), Al (0.24 – 0.37 wt% Al2O3), and Ca (0.75 – 0.94 wt% CaO) and minor depletion in Fe (29.21 – 30.59 wt% Fe2O3) and Na (12.43 – 12.63 wt% Na2O). In Kimb65d, aegirine contains many impurities, with enrichment in Ti (2.32 – 3.01 wt% TiO2), Mg (3.39 – 6.34 wt% MgO), and Ca (9.28 – 14.16 wt% CaO) with minor Mn (0.26 – 0.37 wt% MnO), and depletion of Fe (20.71 – 25.12 wt% Fe2O3), and Na (5.58 – 8.09 wt% Na2O). It should be noted that the xenoliths from which these crystals were analyzed do not allow for the determination of a granitic or gneissic protolith. Without a larger sample size of xenolith reaction minerals across a number of xenoliths, it cannot be ruled out that differences in xenolith mineral chemistry do not result from differences in the mineralogy of individual xenoliths.  Diopside is a country rock xenolith reaction mineral present in the reaction rims of xenoliths, and is also the second most common mineral in the interclast matrix where it forms microlites. No significant difference in composition can be observed for diopside in either textural position; all show enrichment of Al (0.14 – 0.92 wt% Al2O3), Fe (3.03 – 8.06 wt% FeO), Mn (0.14 – 0.28 wt% MnO) and Na (0.57 – 1.59 wt% Na2O). Diopsides hosted within the interclast matrix of Kimb65a show additional enrichment of Ti (0.55 – 2.36 wt% TiO2), Na (1.17 – 4.56 wt% Na2O) and K (0.27 – 1.17 wt% K2O).  Pectolite (Na2Ca2Si3O8(OH)) is a relatively uncommon country rock xenolith reaction mineral in Kimb65b and Kimb65d. Mineral compositions are relatively pure and consistent in both kimberlite units; pectolites show minor enrichment of Al (0.03 – 0.49 wt% Al2O3), Fe (0.16 – 1.00 wt% FeO), Mn (0.09 – 0.24 wt% MnO), Mg (0.16 – 5.01 wt% MgO) and minor depletion in Na (6.49 – 8.92 wt% Na2O). Results - Mineral Chemistry  70  3.2.8 Amphibole Amphiboles are observed only in xenoliths hosted within volcaniclastic rock types of Kimb65a. They crystallize between the grain boundaries of quartz, plagioclase, and alkali feldspar and preferentially replace quartz. Their compositions (Table 3.9) are variable even within a single xenolith but all are Na-rich (4.97 – 8.55 wt% Na2O) and K-rich (2.19 – 7.44 wt% K2O). Ca is not reported as it was not observed in EDS spot analyses and was not analysed in WDS. Even if Ca had been present and analyzed it could not have been greater than 3.5 wt% CaO; thus these are considered to be Na-K amphiboles generally consistent with classification as the rare amphibole eckermannite – NaNa2(Mg4Al)Si8O22(OH)2, which is more alkali-rich and less aluminous than glaucophane – Na2(Mg3Al2)Si8O22(OH)2. The compositions also show significant variation in Al (0.96 – 7.52 wt% Al2O3), Fe (8.50 – 12.21 wt% FeO), and Mg (9.45 – 15.67 wt% MgO) but despite these variations the occupancy of the Y-site by Al, Fe and Mg in the general amphibole formula W0-1X2Y5Z8O22(OH,F)2 is maintained at approximately 5. Table 3.9 – Compositions of amphibole in crustal xenoliths in volcaniclastic rock types in Kimb65a  Renard 65 Rock Classification  71  4. Renard 65 Rock Classification 4.1 Comparison of Features with Group I Kimberlites Kimberlites share a number of petrographic similarities with their related rock types: lamproite, melnoite, alnoite, olivine melilitite, and in particular orangeite. These related rocks are believed to represent compositionally distinct magmas (Mitchell, 1986, 1991, 1994, Mitchell & Bergman 1991), and the accurate identification of a parental melt type is crucial to the interpretation of magma genesis and emplacement. Kimberlites can be distinguished from most of the related rock types (lamproite, melnoite, alnoite, and olivine melilitite) based on their petrographic features and their consistency with the definition of kimberlite as presented in Sect. 1.2 Kimberlite Defined (Wooley et al., 1995). On this basis, the Renard 65 kimberlites which are described as olivine macrocryst rich (5 – 30 % modal abundance) and olivine phenocryst rich (3 – 25 % modal abundance), with groundmass and magmaclast mineral assemblages containing phlogopite, spinel, perovskite, apatite, calcite and serpentine, are petrographically consistent with Group I kimberlites. However, these features are largely consistent with orangeites as well, which only differ petrographically from Group I kimberlites in that they show higher modal abundances of microphenocrystal and groundmass phlogopite, lower modal abundances of groundmass spinel and perovskite, groundmass monticellite is never present, and groundmass diopside is common (Mitchell, 1995). Given these petrographic similarities, it can be difficult to confidently differentiate between Group I kimberlites and orangeites on a petrographic basis alone, and commonly geochemical trends play an important role in parental magma type determination (Smith et al., 1978; Boctor & Boyd, 1982; Skinner & Scott, 1979; Mitchell & Meyer, 1989; Mitchell, 1986, 1991, 1994, 1995; Mitchell & Bergman, 1991; Birkett et al., 2004; van Straaten et al., 2007). Magmatic evolution trends of groundmass spinel (towards magnesian ulvöspinel-ulvöspinel-magnetite solid-solutions) and mica (towards tetraferriphlogopite compositions) and REE enrichment trends in groundmass perovskite, and Sr enrichment trends in groundmass apatite, are all geochemical indicators of orangeite (Mitchell, 1995). Renard 65 Rock Classification  72  The following mineralogical and geochemical features of the Renard 65 kimberlites are consistent with features of Group I kimberlites, based on criteria proposed by Mitchell (1995): (1) Diopside does not occur as a homogeneously distributed primary mineral in the interstitial groundmass or within magmaclasts. (2) Groundmass carbonates are all calcites; varied carbonates that may be present in Group II kimberlites including: norsethite, strontianite, and witherite, are absent. (3) Groundmass perovskites are large (20 – 40 µm) and abundant (5 – 10 %) with limited REE-enrichment (< 4.5 wt % ∑REE). (4) Groundmass spinels commonly display atoll textures, and their compositions are Cr-rich (Cr/(Cr+Al) = 0.75 - 0.85) with more evolved mantles enriched in Ti and Al, and depleted in Cr (figure 3.20), thus displaying compositions consistent with unevolved magmatic Trend 1 spinels.  (5) Groundmass apatites are relatively poor in Sr (<2.5 wt % SrO). (6) K-Ba titanates and zirconium silicates are absent. (7) The macrocryst mineral assemblage is entirely dominated by olivine with very limited phlogopite. The only feature of the Renard 65 kimberlites that is more consistent with orangeite than Group I kimberlite is the compositional trend of groundmass micas; while the majority of groundmass, interclast matrix, and crustal xenolith reaction corona phlogopite plots within the compositional field for Group I kimberlites (Mitchell, 1995) (figure 4.1), phlogopite within magmaclasts and within the interclast matrix of KPKt rock types in Kimb65d vary in composition from phlogopite cores to Al-depleted tetraferriphlogopite rims (figure 4.1, 3.17D & Table 3.4), a trend that is commonly observed in Orangeites (Group II kimberlites) (Mitchell, 1995). Kinoshitalite, the Ba-rich (up to ~ 18% BaO) and Al-depleted variety of phlogopite typical of Group I kimberlites (Mitchell, 1995) are notably absent from these rocks (Table 3.4). The occurrence of tetraferriphlogopite in Group I kimberlites is not unique to Renard 65, and has Renard 65 Rock Classification  73  been described in Group I kimberlites from Namibia and China (Mitchell, 1995), Snap Lake (Mogg et al., 2003) and Gahcho Kue (Caro & Kopylova, 2004). Caro and Kopylova (2004) proposed that this apparent affinity for the orangeite compositional trend in groundmass micas in Group I kimberlites may result from Si contamination by crustal xenoliths; significant variations in phlogopite compositions within a single rock type may result from local diffusion controlled crystallization and may not reflect changes in the bulk composition of the magma. Thus they argue that groundmass mica compositions may not serve as a viable petrogenetic indicator of changing bulk-composition in kimberlites which are contaminated by silicic crustal xenoliths. We similarly propose that since the core compositions of phlogopite in Kimb65d are consistent with Group I kimberlites, and the tetraferriphlogopite rim compositions deviate into the Orangeite field (figure 4.1), that this compositional zoning must reflect late-stage and likely subsolidus processes and do not serve as a petrogenetic indicator of changing bulk composition in the melt.   Figure 4.1 – Al2O3 vs. TiO2 and Al2O3 vs. FeO diagrams for groundmass and magmaclast hosted micas in Renard 65. Fields for Group I kimberlite and Group II Kimberlite (Orangeite) are from Mitchell (1995). The majority of groundmass and magmaclast hosted micas plot within the Group I kimberlite compositional field, while the rims of phlogopites from KPKt rock types in Kimb65d plot within the Group II kimberlite (orangeite) field. Kimb65b refers to HK rock types while Kimb65b* refers to HKt rock types. C = core and R = rim. Note that mica compositions do not deviate by textural position within the same rock type; micas hosted within groundmass, interclast matrix or reaction coronas on crustal xenoliths display comparable compositions within their host rock type and geological unit. Renard 65 Rock Classification  74  The Fe enrichment of phlogopite rims hosted in both magmaclasts and within the interclast matrix of KPKt rock types in Kimb65b (figure 4.1) may result from re-equilibration with subsolidus deuteric fluids. It will be shown in Sect. 5.6.2 Reactions Between Crustal Xenoliths in HK, HKt and KPKt Rock Types in Kimb65b and Kimb65d, that subsolidus interaction of the kimberlite with deuteric fluids (both inherently rich in Mg and enriched in Mg by isovolumetric olivine serpentinization) may result in Mg-Fe exchange in primary spinel, and compositional zoning of diopside to aegirine within reaction coronas on crustal xenoliths; these features were observed to be increasingly developed from HK to HKt to KPKt rock types in Kimb65b and Kimb65d respectively. It is therefore suggested that the development of tetraferriphlogopite rim compositions on phlogopite in KPKt rock types in Kimb65d likely reflect interaction with subsolidus fluids with Fe enrichment resulting from the exchange of Mg for Fe in spinel. Additionally, the fact that this compositional zoning is developed on micas both within the magmaclasts and within the interclast matrix would suggest that like groundmass mica, interclast matrix micas are crystallized prior to interaction with fluids (deuteric or otherwise), and are (at least in Renard 65) highly unlikely to be of hydrothermal origin; this significance of this point will become clear as we delve into the origins of the interclast matrix in Sect. 5.2 Current Models on the Origin of the Interclast Matrix in KPKs. 4.2 Mineralogical Classification of Renard 65 Kimberlites Clement and Skinner (1979) after working on a number of kimberlite pipes in southern Africa recognized that the primary kimberlite mineral assemblages of the microphenocryst, groundmass and mesostasis suites were the most diagnostic components of a kimberlite. They devised a classification scheme in which kimberlites are mineralogically classified based on the modal mineralogy of these interstitial groundmass and magmaclast hosted primary mineral assemblages. The mineralogical classification of kimberlites is an essential step in allowing for a meaningful comparison of the mineral assemblages and mineral chemistry of kimberlite phases in a single pipe, kimberlites in different pipes within a province, or kimberlites in different provinces. The current mineralogical classification of Scott Renard 65 Rock Classification  75  Smith et al., (2013) is based on the principle of Clement and Skinner (1979) and involves the normalization of groundmass modal mineralogy to 100% on an olivine-free basis, with kimberlite names prefixed by the most abundant groundmass mineral, with additional prefixes for any groundmass mineral with a normalized modal abundance > ½ of the most abundant mineral. For example, the normalized modal mineralogy of the magmaclast hosted groundmass in Kimb65a (Table 4.1) shows that phlogopite is the most abundant groundmass mineral (42 % normalized modal abundance), followed by apatite (23 % normalized modal abundance). Since the normalized abundance of apatite is greater than ½ of the normalized abundance of phlogopite, it is included in the mineralogical classification. As a result, Kimb65a kimberlites are mineralogically classified as apatite-phlogopite kimberlite; note that the minerals are  Table 4.1 Mineralogical Classification of Renard 65 Geological Units  Renard 65 Rock Classification  76  listed in order of increasing modal mineralogy. Mineralogical classifications apply to kimberlite for each of the geological units in Renard 65 (Table 4.1); with the exception of serpentine and calcite, the primary interstitial groundmass and magmaclast mineralogy was observed to be consistent between areas of a single thin section, between thin-sections from the same sample, and between samples collected at various depths within each geological unit. These mineralogical classifications can thus be applied to all kimberlites within a given geological unit. 4.3 Textural-Genetic Classification The volcaniclastic kimberlites in the Renard 65 pipe show features that are consistent with their classification as Kimberley-type pyroclastic kimberlites (Scott Smith et al., 2013): 1) The kimberlites contain spherical uncored magmaclasts (figure 3.1A, 3.1D & 3.2A) or thin-selvage magmaclasts cored on olivine macrocrysts and crustal xenoliths (figure 3.1B, 3.15A, 3.15B & 3.15C) that are inconsistent with the fluidal and amoeboid-shaped magmaclasts common in Fort-à-la-Corne type pyroclastic kimberlites (Scott Smith, 2008). 2)  Olivine macrocrysts show (in macrospecimen) a regular size distribution of fine to coarse (1-8 mm), with no evidence of grain-size sorting via pyroclastic sorting or winnowing processes as are common in FPK’s (Scott Smith, 2008a). 3) Olivine macrocrysts are subround to round, and when preserved they do not show deformation lamellae. Angular macrocrysts with deformation textures are consistent with explosive pyroclastic emplacement processes interpreted in FPK’s (Scott Smith, 2008). 4) The kimberlite pipe shows steep-sided walls, emplaced through competent gneissic-granitoid basement rocks and the kimberlites contain a generally very high modal abundance (20-90 %) of subround to angular crustal xenoliths (Scott Smith 2008b).  Renard 65 Rock Classification  77  5) The interclast matrix of the volcaniclastic kimberlites is devoid of carbonate, and contains microlitic assemblages of phlogopite and diopside with interstitial serpentine and irregularly distributed primary kimberlitic groundmass minerals (Hetman, 2004; Mitchell et al., 2009; Scott Smith 2008b). In addition to these features, some of the volcaniclastic kimberlites show transitional textures intermediate between KPK and coherent kimberlite. Kimb65b is composed primarily of coherent kimberlite that varies gradationally on a centimeter to meter scale showing an increase in crustal xenolith abundance and development of a mottled texture defined by patches of microlitic diopside (figure 3.6B, 3.6F & 3.6D). Kimb65a is composed primarily of volcaniclastic kimberlite that varies gradationally showing a decrease in crustal xenolith abundance and an increasingly incipient magmaclastic texture in which the magmaclasts show a higher degree of interconnectivity as well as a higher modal abundance (figure 3.1C). These gradational changes in texture from KPK to KPKt, and HK to HKt in the geological units Kimb65a and Kimb65b are consistent in character with complete HK to KPK transition zones documented in KPK pipes in South Africa (Clement, 1982; Clement & Reid, 1989; Skinner and Marsh, 2004; Mitchell et al., 2009) and Canada (Hetman et al., 2004; Hetman, 2008; Scott Smith 2008b; Armstrong et al., 2008; Kupsch & Armstrong, 2013, Muntener & Scott Smith, 2013). Based on the similarities of kimberlite components, rock textures, and pipe morphologies, the Renard 65 kimberlite is classified as a Kimberley-type pyroclastic kimberlite. Consequently, the rock textures of the various geological units in Renard 65 will now be referred to as HK, HKt, KPKt and KPK in place of coherent, transitional coherent, transitional volcaniclastic and volcaniclastic. Discussion - The Origin and Ascent of Kimberlite Magmas  78  5. Discussion 5.1 The Origin and Ascent of Kimberlite Magmas Kimberlitic melts originate in the sub-cratonic asthenospheric mantle at depths greater than any other known igneous rock-type; they are low-viscosity alkaline, ultrabasic, silica-undersaturated (< 30 wt. % SiO2), and volatile-rich (7-20 wt. % CO2+H2O) (Price et al., 2000; Le Roux et al., 2003; Harris et al., 2004; Becker and Le Roux, 2006; Kopylova et al., 2007; Sparks et al., 2009, Sparks et al., 2006; Kavanagh and Sparks, 2009). Kimberlite magmas ascend from depths of approximately 250 km within vertical fractures (as dykes). Their ascent is propagated by a pressure gradient between the magma-filled dyke (initially ~ 8 GPa) and the lower-pressure dyke-tip (initially ~ 2 GPa), which is filled and continuously recharged with exsolved, CO2-rich, supercritical fluids as a result of decompression during ascent (Wilson & Head, 2007). Magma flow within the dyke is suggested to be turbulent with estimated ascent speeds ranging from 4 to 20 m/s (Sparks et al., 2006; Wilson & Head, 2007). The ascent rate, and the typical dimensions of kimberlite feeder dykes (0.3 – 1.0 m width and 0.5 – 10 km length along strike) after Mitchell, 1986; Sparks et al., 2006) suggest magma supply rates of 500 to 105 m3/s. This in combination with minimum estimates of kimberlite eruption volumes (estimated from volumes of preserved pipe-infill and extra-crater deposits) of 106 to 2 x 108 m3 suggest minimum eruption durations of several hours to several weeks (Sparks et al., 2006), erupting presumably as sub-Plinian to Plinian volcanoes (Sparks et al., 2006).  The solidus temperature at which the kimberlite is fully crystallized after emplacement is poorly constrained because (i) kimberlite melt compositions are not well defined (Price et al., 2000; Le Roux et al., 2003; Harris et al., 2004; Becker and Le Roux, 2006; Kopylova et al., 2007; Sparks et al., 2009, Sparks et al., 2006; Kavanagh and Sparks, 2009), and (ii) no experiments have been performed on the low pressure crystallization of natural kimberlite melt. The low pressure solidus has been modelled by analogy with experiments on the simple synthetic systems CMS-H2O-CO2 and demonstrated to be approximately Discussion - Kimberlite Emplacement: Processes of Diatreme Formation and Infill  79  600 °C at 1 kbar (Frantz & Wyllie, 1967) and 670 °C at 2 kbar (Otto & Wyllie, 1993). As summarized by Mitchell (1986), the primary liquidus phases in kimberlite melts are phenocrysts and microphenocrysts with subhedral to euhedral crystal habit, or minerals which can otherwise be shown to have crystallized in-situ to form the interstitial groundmass of the kimberlite; the primary liquidus phenocrystic phases include olivine, phlogopite and chromite, and the primary liquidus microphenocrystic phases include olivine, phlogopite, Ti-bearing spinels, perovskite, ilmenite, diopside, monticellite, apatite, calcite and serpentine.  5.2 Kimberlite Emplacement: Processes of Diatreme Formation and Infill The nature of the kimberlite diatreme forming and infilling processes has been an ongoing subject of debate over the past few decades. There are currently two emplacement models that broadly overlap in their initial diatreme formation processes, but diverge in (i) the extent of diatreme excavation during explosive eruption, (ii) the process by which the diatreme is infilled, and (iii) the application of the process referred to as ‘fluidization’ as it pertains to the formation of magmaclasts. The first emplacement model which will here be referred to as the ‘subvolcanic fluidization’ model, is an emplacement model that envisions kimberlite diatreme formation caused by subvolcanic fragmentation and authigenic brecciation of the host rock. As critical pressures build in the CO2-rich fluid-filled cavity at the dyke-tip (as described in the previous sub-section) explosive breakthrough is triggered causing rapid decompression of the magma. The competent nature of the host country rock allows for initial subvolcanic fragmentation and degassing resulting in a downward-propagating degassing front in the magma column, with exsolved magmatic volatiles streaming upwards. This rapidly exsolving magmatic volatile phase is proposed to induce a state of fluidization in the above magma column, whereby the proportion of the exsolved fluid phase exceeds 70 % causing the magma to fragment into droplets (molten magmaclasts) suspended in fluid; this process would be most vigorous at the top of the magma column, Discussion - Kimberlite Emplacement: Processes of Diatreme Formation and Infill  80  with decreasing proportions of exsolved fluids with depth. Concomitantly, the turbulently fluidized magma causes authigenic brecciation and entrainment of the wall rock, carving out the steep-sided diatreme. Rapid cooling of the magma during this short-lived degassing and fluidization stage results in ‘freezing’ of the magma preserving the in-situ magmatic textural gradations from coherent kimberlites in the root zone to transitional and pyroclastic kimberlites in the middle to upper diatreme zone. This model was first proposed by Clement (1982), and later re-summarized by Field & Scott Smith (1999), Skinner & Marsh (2006), and Skinner (2008). A very similar model was also proposed by Wilson & Head (2007). The use of the term fluidization originates from industrial fluidization processes where a high velocity gas is fluxed through a fixed bed of fine-particles at sufficient velocity to induce bed fluidization (fluid-like behavior); beds containing mixed particle sizes become homogenized, and accretionary spherules form by agglutination (McCallum, 1985; Gernon et al. 2012). These industrial experiments suggested that similar processes may explain the homogenization of diatreme infilling material, and the formation of spherical magmaclasts (Field & Scott Smith, 1998, 1999; Scott Smith, 2008; Skinner, 2008; Gernon et al., 2012).  This model gained significant traction after Hetman et al. (2004) documented gradational changes in rock texture with depth from KPK rock types, to KPKt and HKt transitional rock types, down to HK rock types in the root zones of the Gahcho Kue kimberlites. These observations served as a significant pillar of support for the subvolcanic fluidization emplacement model as these preserved gradational rock types were argued to represent an upwelling frozen degassing front (Hetman et al., 2004; Scott Smith, 2008; Skinner, 2008). While this model does provide adequate explanation for the observed features in these deposits, fluidization of this style is not a process that has been demonstrated in any other volcanic system and the suggestion that this process should be unique to kimberlites has been met with criticism (Stripp et al., Discussion - Kimberlite Emplacement: Processes of Diatreme Formation and Infill  81  2006; Sparks et al., 2006; Walters et al., 2006; Cas et al., 2008; Hayman et al., 2008, 2009; Buse & Sparks, 2010). The second emplacement model which will here be referred to as the ‘volcanic fragmentation’ model, is an emplacement model that envisions kimberlite diatreme formation caused by rock-bursting, undermining, down-faulting, and crater-rim slumping causing the diatreme to grow in both depth and width during sustained explosive volcanic eruption. According to this model, kimberlite volcanoes behave like regular sub-Plinian to Plinian volcanoes. During the initial explosive eruption stage, significant volumes of wall-rock are ejected from the pipe and thus this stage is largely one of erosion and pipe excavation. At a certain threshold where the diatreme has grown beyond a critical cross-sectional area, the gas exit velocities decline rapidly and the diatreme is infilled by (i) extra-crater material deposited by debris flows, (ii) pyroclastic material deposited from the collapsing eruption column, and (iii) wall-rock material deposited by rock-bursting, undermining, and slumping (Sparks et al., 2006; Walters et al., 2006; Porritt & Cas, 2009; Hayman & Cas, 2011; Gernon et al., 2012).  There exists some disagreement between these authors though on the role of fluidization (or absence of fluidization) during the waning stages of eruption. Some authors suggest that the majority of fine-ash particles are elutriated during the explosive volcanic eruption and as a result the mixture of pyroclastic and slumped material infilling the diatreme is largely (30 %) porous. This high porosity would allow the tephra to become fluidized by the continued gas exit flows from the underlying degassing magma, causing lithic clasts and xenocrysts to become coated by the degassing kimberlite melt by a process referred to as ‘fluidized spray granulation’ thus producing spherical magmaclasts cored on lithic clasts (Sparks et al., 2006; Walters et al., 2006; Gernon et al., 2012). Other authors reject this and in-fact any role of fluidization on the basis that pyroclastic kimberlites often display matrix-supported textures indicating fine-particle elutriation was minor and that the interstices between juvenile pyroclasts and lithic Discussion - Kimberlite Emplacement: Processes of Diatreme Formation and Infill  82  clasts was likely volcanic ash (Hayman et al., 2009; Porritt & Cas, 2009; Hayman & Cas, 2011); though, they do not offer any alternative explanation for the formation of magmaclasts. All proponents of this volcanic fragmentation emplacement model agree though that late-magmatic to post-magmatic alteration plays a significant role in overprinting the primary textures of pyroclastic kimberlites.  It is important to note that the first emplacement model is an attempt to describe the emplacement processes that account for the features of specifically Kimberley-type pyroclastic kimberlite pipes, while the second model and variations thereof are based on single deposits that are not consistent with the features of Kimberley-type pyroclastic kimberlites, and address features that have been described solely in resedimented volcaniclastic kimberlites and Fort-à-la-Corne-type pyroclastic kimberlites. The volcanic fragmentation emplacement model is demonstrated to be inapplicable in describing the emplacement processes of Kimberley-type pyroclastic kimberlites. Firstly, Sparks et al. (2006) propose that during the diatreme formation stage of emplacement the pipe is (until some point in time) largely excavated, supporting this proposition with the observation that some kimberlite pipes (Ekati, Diavik and Lac de Gras pipes in the NWT) contain surficial material (i.e. tree-bark), representing a case where the diatreme must have been largely empty and infilled by crater-wall and pipe margin slumps. Similarly, Hayman & Cas (2011) propose that the diatreme of the Jericho kimberlite was likely fully excavated and infilled by rock bursting and high-density debris flow deposits based on the identification of variably sorted interstratified breccia lenses within the diatreme. As pointed out by Scott Smith (2008), the presence of surficial debris and bedded and or sorted material have never been observed in Kimberley-type pyroclastic kimberlites, but only in resedimented volcaniclastic kimberlites which display entirely different megascopic structures, textures and mineralogy. This evidence cannot be used to support such processes in the formation of KPK pipes. Scott Smith (2008) further points out that the presence of strongly juxtaposed internal contacts between units, steeply oriented fabrics, Discussion - Kimberlite Emplacement: Processes of Diatreme Formation and Infill  83  and the presence of abundant and large locally derived country rocks are inconsistent with the notion that the diatreme was ever fully excavated.  Secondly, Sparks et al. (2006) and Gernon et al. (2012) suggest that sustained high temperatures and high magma resupply rates may have resulted in particle sintering and agglutination to produce the observed textural gradations from HK to KPK rock types in KPK pipes (as identified at the Gahcho Kue kimberlites by Hetman et al., 2004). They point to earlier work by Brown et al. (2008) on the Venetia K2 pipe and the Orapa B/K9 pipes which showed that geological units previously described as coherent kimberlite displayed evidence of pyroclastic welding including (i) sintering and deformation between magmaclasts and, (ii) the formation of “calcite domains in the interstices between “small plastically deformed juvenile ash grains and groundmass coated pseudomorphs of olivine phenocrysts” (Brown et al., 2008). It is here crucially important to point out that while these features are very difficult if not impossible to identify in drill core they are easily identifiable in thin-section as shown by Brown et al. (2008). These pyroclastic welding textures and the presence of “interstitial calcite domains” are completely and entirely uncharacteristic of KPK rock types, though they are arguably characteristic of Fort-à-la-Corne type pyroclastic kimberlites. It is thus unreasonable to extrapolate these processes to explain preserved transition zones in the Gahcho Kue kimberlites which display entirely different mineralogical and textural characteristics for which detailed petrography was publically available. Furthermore, independent data from Scott Smith (2008) on the Fort-à-la-Corne pyroclastic kimberlite demonstrates that while in most massive extrusive volcaniclastic kimberlites olivine populations are highly variable and different from the size distributions observed in HK rock types as a consequence of eruption related processes (i.e. sorting/winnowing). The olivine size distributions in KPK rock types however, are consistent with those of HK rock types, further substantiating the subvolcanic fluidization emplacement model whereby KPK rock types are the result of in-situ magmatic textural Discussion - Current Models on the Origin of the Interclast Matrix in KPKs  84  modification of the kimberlite magma. This is in contrast to kimberlite descriptions from Jericho, Venetia and Orapa from authors of the volcanic fragmentation model (or authors who utilize this model) who describe various degrees of sorting and crystal enrichment as a result of fine-particle elutriation (Sparks et al., 2006; Stripp et al., 2006; Hayman & Cas, 2011).  We thus conclude that the subvolcanic fluidization model is the only emplacement model that is able to account for the features of Kimberley-type pyroclastic kimberlites and is the model that will be utilized and modified in the interpretation of the Renard 65 kimberlites as well as other Kimberley-type pyroclastic kimberlites. 5.3 Current Models on the Origin of the Interclast Matrix in KPKs The most direct way of constraining the emplacement-related processes that produce KPK rock types is to address the features that distinguish them from HK rock types. Among many differences perhaps the most pronounced is the presence of an interstitial microlitic matrix (commonly: diopside + phlogopite + serpentine) in place of a crystalline interstitial groundmass (commonly: monticellite + phlogopite + serpentine + carbonate). Other notable features that distinguish KPK rock types include the presence of magmaclasts (formerly referred to as pelletal lapilli), high abundances of weakly reacted crustal xenoliths, and low abundances of strongly serpentinized olivine. The development of emplacement models for KPKs has been based largely on explaining and attributing the origin of the interclast matrix mineral assemblages to either (i) crystallization products of the exsolved vapour phase and deuteric fluids from a residual volatile-rich kimberlite melt (Clement, 1982; Mitchell, 1986; Field & Scott Smith, 1999; Skinner & Marsh, 2004), or (ii) progressive crystallization (either into porous void space or replacing volcanic ash) resulting from interaction between hydrothermal fluids, olivine, and silicic crustal xenoliths (Stripp et al., 2006; Sparks et al., 2006; Walters et al., 2006; Hayman et al., 2008, 2009; Buse & Sparks, 2010). The important difference between them being that in the first model the textures and mineralogy of the rock Discussion - Current Models on the Origin of the Interclast Matrix in KPKs  85  types are considered to be primary and thus provide insight into the emplacement processes of the kimberlite magma, while the second model considers the textures and mineralogy to be secondary in nature thus obscuring the original emplacement process of the kimberlite magma. Here we will examine these two models for the origin of the interclast matrix assemblage in KPKs and determine their viability in accounting for the features of both the Renard 65 kimberlites and other KPK pipes. The first model which accompanies the subvolcanic fluidization emplacement model, envisions that as a volatile-saturated kimberlite magma reaches surface the initial explosive breakthrough and reduction in confining pressure will trigger (i) a catastrophic exsolution of magmatic volatiles into a subcritical vapour phase, and (ii) authigenic brecciation of the host country rock; the exsolving magmatic volatiles will stream up through the magma column expanding adiabatically causing the magma to become fluidized, fragmenting into magmaclasts which become spherical by surface tension processes. The remaining vapour phase quenches to precipitate common interclast matrix minerals (including diopside, phlogopite and serpentine) by shifting the stable mineral assemblage from monticellite + phlogopite + calcite + serpentine to diopside + serpentine + phlogopite by exsolution of CO2 (Field & Scott Smith, 1998, 1999; Skinner & Marsh, 2004) as described by reaction (1). KMg3AlSi3O10(OH)2   +   CaCO3   +   Mg3Si2O5(OH)4   +   2SiO2   +   MgO →                           (1) CaMgSi2O6   +   Mg3Si2O5(OH)4   +   KMg3AlSi3O10(OH)2   +   2CO2 phl   +   cal   +   srp   +   SiO2   +   MgO   →   di   +   srp   +   phl   +   CO2 This model gained significant traction after Hetman et al. (2004) documented gradational changes in rock texture with depth from KPK rock types, to KPKt and HKt transitional rock types, down to HK rock types in the root zones of the Gahcho Kue kimberlites. These observations served as a significant pillar of support for the subvolcanic fluidization emplacement model as these preserved gradational rock types were argued to represent an upwelling frozen degassing front (Hetman et al., 2004; Scott Smith, 2008; Skinner, 2008). Interestingly, despite the fact that increased felsic xenolith abundances have been correlated with Discussion - Current Models on the Origin of the Interclast Matrix in KPKs  86  increasingly transitional to KPK rock types both at Gahcho Kue as well as other KPK occurrences (Hetman et al., 2004; Muntener and Scott Smith, 2013), the role of xenolith assimilation, in particular the assimilation of Si4+ into the kimberlite melt and the effects that may have in driving reaction (1) were not explored or even suggested by proponents of this model. This will become a significant topic of discussion to follow in Sect. 5.8 Effects of Crustal Xenolith Assimilation on Volatile Fluid Exsolution in Kimberlite Melt. While this model does provide adequate explanation for the observed features in these deposits, fluidization of this style is not a process that has been demonstrated in any other volcanic system and the suggestion that this process should be unique to kimberlites has been met with criticism (Stripp et al., 2006; Sparks et al., 2006; Walters et al., 2006; Cas et al., 2008; Hayman et al., 2008, 2009; Buse & Sparks, 2010). The second model which accompanies the volcanic fluidization emplacement model, contends that more a more conventional process such as air-fall deposition of pyroclastic components from an eruption column should be presumed as the dominant pipe-infilling process, and that a substantial post-emplacement meteoric fluid flux may have produced the low temperature interclast matrix assemblage of diopside and serpentine ± brucite or magnetite as a product of hydrothermal metamorphism (Stripp et al., 2006; Sparks et al., 2006; Walters et al., 2006; Hayman et al., 2008, 2009; Buse & Sparks, 2010; Gernon et al., 2012). Some proponents of the hydrothermal model argue that meteoric water infiltrating highly porous tephra will serpentinize olivine in an isovolumetric reaction (2) releasing Mg2+ and Si4+ into the serpentinizing fluids; at the same time, felsic crustal xenoliths containing plagioclase (i.e. dolerite) are replaced by serpentine releasing Ca2+ into the fluid (Stripp et al., 2006; Sparks et al., 2006; Walters et al., 2006; Buse & Sparks, 2010). Additional serpentine would then be able to precipitate into the interstitial void-space in the tephra deposit by consuming Mg2+ and Si4+ derived from olivine serpentinization, and additional Si4+ derived from the breakdown of crustal xenoliths by reaction (3). Diopside would crystallize as any or all of (i) radial fibrous microlites around microcrysts by reaction between Ca-bearing fluids and Discussion - Current Models on the Origin of the Interclast Matrix in KPKs  87  talc rims on microcrysts by reaction (4), (ii) direct precipitation from the hydrothermal fluids in the interclast matrix by reaction (5), or (iii) crystallization at the interface between calcite and serpentine in the matrix by reaction (6). These five reactions were proposed by Stripp et al. (2006); the notations “ol” and “cr” in reaction (3) are reference to components in aqueous solution derived from the breakdown of olivine or crustal xenoliths. 2.000Mg2SiO4 + 1.613H2O(aq) → 0.806Mg3Si2O5(OH)4 + 1.581MgO(aq) + 0.387SiO2(aq)                (2) olivine            H2O(aq)                 serpentine                  MgO(aq)                SiO2(aq)  1.54MgO + 0.36SiO2(aq) + 0.66SiO2(aq) + H2O → 0.51Mg3Si2O5(OH)4                             (3)                                        MgO(aq,ol)      SiO2(aq,ol)            SiO2(aq, cr)      H2O           serpentine Mg3Si4O10(OH)2 + Ca(OH)2(aq) → CaMgSi2O6 + 2Mg(OH)2 + SiO2(aq)                            (4)                                              talc               Ca(OH)2(aq)           diopside        brucite       SiO2(aq) 2SiO2(aq) + CaO(aq) + MgO(aq) → CaMgSi2O6                                              (5)                       SiO2(aq)      CaO(aq)    MgO(aq)       diopside Mg3Si2O5(OH)4 + 3CaCO3 + 4SiO2(aq) → 3CaMgSi2O6 + 3CO2 + 2H2O                       (6)                             serpentine        calcite      SiO2(aq)           diopside        CO2       H2O Other proponents of the hydrothermal model argue that the commonly observed matrix-supported textures of KPK rock types is evidence that the initial deposit was not a porous clast-supported and fine-particle elutriated tephra, but more likely a variably clast- to matrix-supported tephra with interstitial volcanic ash (Hayman et al., 2008, 2009; Cas et al., 2008). These authors generally seem to agree that the meteoric fluids must serpentinize olivine and silicic crustal xenoliths to acquire Mg2+ and Si4+, but suggest that Ca2+ may either be derived from crustal xenoliths or the interstitial volcanic ash to crystallize microlitic diopside. They attribute the observed interclast matrix assemblages of serpentine + chlorite + saponite + monticellite + diopside + phlogopite to be the result of overprinting of an originally clastic matrix.  These hydrothermal models are considered by volcanologists to reflect a more realistic scenario in which Kimberley-type pyroclastic kimberlites are emplaced by conventional pyroclastic depositional processes from an eruption column and the unique features of KPKs are the result of hydrothermal Discussion - Current Models on the Origin of the Interclast Matrix in KPKs  88  alteration. There are however, some significant problems in extrapolating these hydrothermal models that are based on single localities to account for the characteristic aspects of KPK deposits internationally: 1) Stripp et al. (2006) described MVK in the Venetia kimberlite as variably to extensively altered, whereas silicic crustal xenoliths in KPK rock types at other localities including Gahcho Kue and Renard (Hetman et al., 2004; Muntener & Scott Smith, 2013) are often described as fresh or relatively unreacted, generally maintaining their original textures and mineralogy; they never display metasomatic rims when hosted in KPK rock types. These extensively altered xenoliths as described by Stripp et al. (2006) are therefore not considered to be representative of the characteristic appearance of KPK rock types at any other locality. 2) Buse & Sparks (2010) describe metasomatic rims on crustal xenoliths in the Orapa B/K9 kimberlites which contain bultfonteinite and hydrogarnet inclusions which they use to constrain the conditions of alteration to low temperatures involving hydrous fluid compositions (see detailed summary in Sect. 5.5 Reactions Involving Crustal Xenoliths and Host Kimberlite in KPKs for more information). They argue that the hydrous nature of the involved fluids precludes alteration by kimberlitic deuteric fluids and thus they conclude that the fluids were meteoric. We acknowledge that the interpreted conditions of reaction of xenoliths in the B/K9 kimberlites are difficult to dispute, but point out that such interpretations cannot be extrapolated and applied to all KPK kimberlites. Indeed, the crustal xenoliths in Renard and Gahcho Kue also display metasomatic rims, but the compositions of these rims are not constrained either to low temperatures or to anhydrous fluid compositions. In-fact (at least in the case of Renard 65) the mineral assemblages of the metasomatic rims on crustal xenoliths show parallels to fenite aureoles in the host rocks around carbonatite intrusions which are well constrained to result from metasomatism by alkali-rich fluids exsolved by the carbonatite magma during emplacement (discussed in more detail in Sect. 5.6 Crustal Xenoliths in Renard 65). Discussion - Current Models on the Origin of the Interclast Matrix in KPKs  89  3) In order to explain matrix assemblages containing diopside and serpentine as having a hydrothermal origin the required activities of Mg2+, Si4+ and Ca2+ in the hydrothermal fluids must all be derived from serpentinization reactions involving olivine, felsic xenoliths and/or volcanic ash. However, sufficiently high activities of these components would be intrinsic in a residual kimberlite melt and associated deuteric fluids. 4) Arguments necessitating the involvement of meteoric fluids are primarily based on the high H2O contents that would be required to fully serpentinize a hypabyssal kimberlite containing approximately 50% modal abundance of olivine ~ 5 – 10 wt% H2O (Mitchell, 2005) as being uncharacteristic of kimberlite melts. Stripp et al. (2009) argue based on their synthetic melt crystallization experiments that such a quantity of H2O far exceeds the low pressure solubility of H2O in either silicate or carbonate melts at 100 MPa. However, more recent synthetic melt crystallization of silica undersaturated melts by Moussallam et al. (2015, 2016) show that the H2O-CO2 dependency at 100 MPa would allow for solubility of up to 5 wt% H2O with 5 wt% CO2. To add to this point, it is not logical to extrapolate this value of 5 – 10 wt% H2O as a required minimum H2O content for all kimberlite melts. KPK rock types have consistently lower total olivine modal abundances (i.e. 7 – 20 % total olivine in KPK rock types in Kimb65a and 20 – 30 % total olivine in KPKt rock types in Kimb65d) (Table 3.1), and olivine in HK rock types in KPK pipes often contains significant quantities of fresh olivine (i.e. up to 30% unserpentinized olivine for HK rock types in Kimb65b) (figure 3.6A). The quantity of H2O required to produce the above features is thus lower than 5 – 10 wt% H2O. For the above reasons it is argued that a hydrothermal origin for the interclast matrix of KPK rock types is unlikely as it is unable to account for the characteristic features of KPK occurrences around the world. The unique features of KPKs are thus proposed to be truly a reflection of emplacement-related processes and have not (in the majority of cases) been significantly modified, obscured or erased by Discussion - Macroscopic Features of Crustal Xenoliths in KPKs  90  hydrothermal alteration. It is suggested that the development of hallmark KPK textures are primarily driven by the exsolution of magmatic volatiles in response to decompression, and as we will explore in this research, by the assimilation of silicic crustal xenoliths. The results of this study for the first time emphasize the importance of this assimilation in defining the mineralogy of the interclast matrix. The objective over the following sections will be to characterize the reactions between crustal xenoliths in their host kimberlite in Renard 65, demonstrate the potential for contamination of the kimberlite melt and resulting effects on the stable mineral assemblages of the kimberlite melt, and to examine the possible effects of melt contamination by xenoliths on the crystallization of common interclast matrix mineral assemblages. It will be shown that (i) crustal xenolith assimilation can modify the stable groundmass mineral assemblage typical of coherent kimberlites and can account for the development of transitional KPK textures, and (ii) that subsolidus processes may vary spatially within a single phase of kimberlite resulting in mineralogical and chemical changes in reaction coronas on xenoliths and adjacent groundmass spinels with implications for the geological modelling of KPK pipes. 5.4 Macroscopic Features of Crustal Xenoliths in KPKs The distribution of crustal xenoliths in KPK pipes tends to range from very high modal abundances (30 – 95 %) in KPK and KPKt rock types characteristic of the diatreme and upper transition zone facies, to very low modal abundances (1 – 30 %) in HKt and HK rock types characteristic of the lower transition zone and root zone facies (Skinner & Clement, 1979; Mitchell 1986; Dawson, 1980; Field & Scott Smith, 1998; Skinner & Marsh, 2004; Hetman et al., 2004; Muntener & Scott Smith, 2013); they also tend to be more angular, and visibly weakly reacted in KPK and KPKt rock types while xenoliths in HKt and HK rock types tend to be more round and visibly strongly reacted (Field & Scott Smith, 1998; Hetman et al., 2004; Muntener & Scott Smith, 2013; Scott Smith et al., 2013). Muntener and Scott Smith (2013) examined over 14,000 m of drill-core and 1000 petrographic thin-sections from the Renard 3 pipe, and characterized all of the kimberlite component features (in much the same manner as Table 3.1) including crustal xenoliths Discussion - Reactions Involving Crustal Xenoliths and Host Kimberlite in KPKs  91  for each geological unit in the pipe; geological units were determined based on primary groundmass assemblages, olivine macrocryst size distributions, microdiamond analysis and indicator mineral chemistry. Their results regarding crustal xenoliths showed that the style of crustal xenolith reaction is generally consistent within and thus characteristic of a geological unit and of a phase of kimberlite for phases containing a narrow range of rock textures. Unfortunately, the xenolith descriptions of Muntener and Scott Smith (2013), much like the style of descriptions carried out in the vast majority of economic geology studies do not identify reaction mineral assemblages, rather referring to the styles of reaction by their diagnostic textures and colours. While such studies provide useful quantitative information on the associations between xenolith features and host rock types, they provide no indication of the processes involved in the interaction between the crustal xenoliths and their host kimberlite. 5.5 Reactions Involving Crustal Xenoliths and Host Kimberlite in KPKs There are only a small handful of detailed studies involving the chronicling of xenolith reaction mineral assemblages and their relationship to emplacement or post-emplacement processes. One of the first detailed mineralogical descriptions of reacted crustal xenoliths in kimberlite comes from Scott Smith et al. (1983), in which ‘kimberlitized’ xenoliths in root zone HKs from South African kimberlites are described as “patches of radiating pectolite, abundant phlogopite, carbonate and serpentine”. They speculated that the occurrence of pectolite as a hybrid mineral both in the reacted crustal xenoliths and in the kimberlite groundmass was a result of contamination of the magma by crustal xenoliths increasing the activities of Na and Si. Along a similar line of thought Caro and Kopylova (2004) described root zone HK rock types in the 5034 pipe of Gahcho Kue, Canada, where granitic xenoliths display reaction coronas consisting of an outer rim of phlogopite and an inner rim of diopside. They attribute the occurrence of diopside as a primary phase to high temperature crystallization from a contaminated hybrid melt. They also showed that the composition of diopside in the groundmass and in the reaction coronas was consistent, thus providing direct evidence that the occurrence of primary diopside in kimberlite groundmass is attributable Discussion - Reactions Involving Crustal Xenoliths and Host Kimberlite in KPKs  92  not just to an unusually high activity of Si but directly to the assimilation of crustal xenoliths. While these studies focus on the magmatic interaction between hot kimberlite melt and entrained cold crustal xenoliths, there has been recent compelling evidence to show that metasomatism between crustal xenoliths and their solidified host kimberlite in some pipes occurs in a subsolidus environment involving a hydrothermal meteoric fluid.  Buse et al. (2010) describe metasomatised basalt xenoliths in a large body of pipe-infilling coherent kimberlite in the B/K9 Kimberlite at Damtshaa, Botswana, which they describe as a partially welded pyroclastic kimberlite based on descriptions from Brown et al. (2008) and Field et al. (2008). The basalt xenoliths are described by Buse et al. (2010) as being typically < 10 cm in size and 2 – 7 % in modal abundance. They vary in appearance from fresh to very strongly altered, with textural evidence suggesting bultfonteinite and chlorite replaced the original augite-plagioclase xenolith assemblage as described by reactions (7 & 8), partially preserving the original ophitic texture. Reactions 7 through 9 are from Buse et al. (2010), note that augite is simplified as diopside in reaction (8). 0.42Na0.4Ca0.6Al1.6Si2.4O8 + 1.75CaO(aq) + 2H2O → Ca2Si2(OH)4 + 0.33Al2O3(aq) + 0.17Na2O(aq)          (7)                                           plag                        CaO(aq)          H2O            bult               Al2O3(aq)             Na2O(aq) 2CaMgSi2O6 + 0.3Al2O3(aq) + 3.6H2O → Ca2Si2(OH)4 + 0.45Mg5Al1.5Si3.5O10(OH)8 + 1.6SiO2(aq)          (8)                                di               Al2O3(aq)         H2O               bult                               chl                             SiO2(aq)                   The xenoliths are rimmed by a metasomatic fringe composed of serpentine with ultra-fine-grained hydrogrossular which are proposed to have overprinted the earlier bultfonteinite and chlorite assemblages at lower temperature as described by reaction (9).  Ca2Si2(OH)4 + 2Mg5Al1.5Si3.5O10(OH)8 → 0.67Ca3Al2Si2O8(OH)4 + 3.33Mg3Si2O5(OH)4 + 0.83Al2O3(aq) + 2H2O (9)             bult                        chl                                      hgr                              serp                             Al2O3(aq)               H2O Buse et al. (2010) showed that the stability of bultfonteinite could be constrained within a temperature range of 250 – 350 °C in the presence of a hydrous fluid (XCO2 < 0.05) based on the upper thermal stability Discussion - Crustal Xenoliths in Renard 65  93  limit of bultfonteinite (600 °C) and the thermal stability of the diopside + calcite matrix assemblage (< 390 C with XCO2 = 0.2) in the serpentinized kimberlite. They also showed that the hydrogarnet compositions within the outer metasomatic serpentine corona could be constrained to crystallization temperatures between 250 – 300 °C. Based on these combined thermal and fluid constraints they propose that the alteration of crustal xenoliths in the welded pyroclastic kimberlite occurred at subsolidus temperatures, with infiltration metasomatism driven by chemical potential gradients in Mg2+ from the serpentinized kimberlite and Si4+ in the basalt xenoliths. The authors propose that the subsolidus reaction temperatures, and the low (<0.05) XCO2 requirement for the fluid phase precludes a deuteric fluid origin. These results are consistent with work by Stripp et al. (2006) showing that the phase equilibria of the assemblage serpentine + diopside in pyroclastic kimberlites at Venetia, South Africa is constrained to T < 380 °C and XCO2 < 0.05. The characterization of common KPK type interclast matrix assemblages as hydrothermally derived is supported by the occurrence of diopside + phlogopite + serpentine assemblages in calcareous mafic skarns, the equilibrium conditions of which can be constrained to 250 – 400 °C (Nabelek et al., 2013).  We examine the interactions between crustal xenoliths and their host kimberlites in Renard 65 assuming the viability of both (i) high temperature in-situ magmatic, and (ii) subsolidus autometasomatic and/or metasomatic reactions. Thermal, fluid and geological constraints are applied in determining the style of interaction during and/or after emplacement. 5.6 Crustal Xenoliths in Renard 65 Two styles of reaction can be recognized among all textural varieties of pipe infilling HK, HKt, KPKt, and KPK rock types in Kimb65a, Kimb65b and Kimb65d in the Renard 65 pipe. Crustal xenoliths in KPK and KPKt rock types in Kimb65a appear relatively unreacted, preserving their original textures and mineralogy and displaying no reaction coronas between the clast margins and the host kimberlite. In contrast, crustal xenoliths in HK, HKt, and KPKt rock types in Kimb65b and Kimb65d appear moderately to very strongly Discussion - Crustal Xenoliths in Renard 65  94  reacted often displaying sequential zonal monomineralic reaction coronas between the clasts and the host kimberlite. These two styles of reaction are described separately in the following two sub-sections. 5.6.1 Reactions between Crustal Xenoliths in KPK rock types in Kimb65a Like similar felsic xenoliths in other KPK and KPKt textural rock types, those in Kimb65a could be described in hand-sample as “fresh” or “relatively unreacted” (Hetman et al., 2004; Muntener & Scott Smith, 2013). However, taking a closer look the xenoliths often appear slightly discoloured green and yellow-beige in an apparent low-degree reaction generally affecting the margins of clasts and penetrating into the core to varying extents based on clast size (figure 3.1B & 3.1E, 5.1). Thin-section petrography, EDS and WDS analyses indicate that the reaction between crustal xenoliths and KPK and KPKt rock types in Kimb65a is characterized by two separate reactions, (i) the high temperature crystallization of eckermannite at the expense of quartz, orthoclase, oligoclase and biotite, and (ii) the subsolidus chloritization of biotite with intergrowths of pyrite.   Figure 5.1 – Polished slab photograph illustrating the typical gneiss and granitoid crustal xenolith assemblage in KPK and KPKt rock types in Kimb65a. The xenoliths are subangular and their primary mineralogy and textures are well preserved. Within the outer ~ ½ cm margins of larger clasts, the clast rims appear discoloured pale green-yellow. For smaller ~1 cm clasts, this discolouration affects the entire clast. The suite of crustal xenoliths thus displays an apparent low-degree reaction.   Discussion - Crustal Xenoliths in Renard 65  95    Figure 5.2 – (A) BSE image showing the occurrence of eckermannite along the margins of quartz, feldspar and biotite. Progressive crystallization appears to be primarily at the expense of quartz as (B) remnant cores of former quartz grains are now mantled by a thin rim of eckermannite and chlorite.   The first reaction that characterizes crustal xenoliths in KPK and KPKt rock types in Kimb65a involves the crystallization of eckermannite, a rare Na-amphibole, by reaction with quartz, oligoclase, orthoclase and biotite. Eckermannite crystallizes between the crystal margins of quartz, feldspars, and biotite (figure 5.2A), and appears to preferentially crystallize at the expense of quartz (figure 5.2B). Synthetic crystallization experiments suggest that at low pressures (1 – 2 kbar) eckermannite is stable over a temperature range of 800 – 1000 °C (Ernst, 1968; Raudsepp et al., 1991). The replacement of quartz, feldspar and biotite to crystallize eckermannite is therefore suggested to have occurred at high temperature (> 800 °C) as described the by the following general reaction (reaction 10). qtz   +   olg   +   or   +   bt   +   H2O   +   MgO(aq)   +   K2O(aq)   +   Na2O(aq)   →   eck           (10) The second reaction that characterizes crustal xenoliths in the KPK and KPKt rock types of Kimb65a is the chloritization of biotite. Biotite crystals are pseudomorphed by chlorite with intergrowths of pyrite (figure 5.3 A & 5.3B) throughout smaller (< 1 cm xenoliths) and within the (~1 cm) margins of larger xenoliths which is what primarily gives the xenoliths their greenish appearance (figure 5.1). The Discussion - Crustal Xenoliths in Renard 65  96  replacement is described by reaction (11) in which biotite reacts with a fluid and is replaced by chlorite with intergrowths of pyrite and releasing SiO2 and K+ into the fluid.  2K(Mg,Fe)3AlSi3O10(OH)2 + 2H2S(aq) → Mg5Al2Si3O10(OH)8 + 2K+(aq) + FeS2 + 3SiO2                (11)      bt                       H2S(aq)                   chl                    K+(aq)       py         qtz This style of reaction suggests temperatures in the range of 150 – 350 °C based on the full range of chloritization temperatures after biotite determined by Yuguchi et al. (2015) and temperatures of lower greenschist facies metamorphism where pyrite is a ubiquitous accessory mineral associated with chloritization (Humphris, 1978). Such occurrences were documented by Li et al. (1998) who described pyrite precipitated as elongated crystals along the cleavage of biotite pseudomorphed by magnesian chlorite for different forms of introduced S as described for example by reaction (12). 4KMgFe2AlSi3O10(OH)2 + 4H2S(aq) + O2 + 4H+(aq) → 2Mg2Fe3Al2Si3O10(OH)8 + 2FeS2 + 6SiO2(aq) + 2H2O + 4K+(aq)(12)               biotite                    H2S(aq)     O2     H+(aq)                          chl                        py        SiO2(aq)       H2O      K+(aq)  Figure 5.3 – (A) Chloritization of biotite in crustal xenoliths begins primarily along the crystal margins and cleavage planes of the biotite crystals, with the total degree of chloritization being quite variable from clast to clast. (B) chlorite pseudomorphing biotite contains common intergrowths of pyrite.  Discussion - Crustal Xenoliths in Renard 65  97  The reactions between crustal xenoliths and their host kimberlite in KPK and KPKt rock types in Kimb65a were likely initiated in a magmatic environment (< 800 °C) as indicated by the crystallization of eckermannite. The entrainment of 40 – 90 % modal abundance of crustal xenoliths at an ambient temperature (< 100 °C) at approximately 5 km depth (Hasterok & Chapman, 2011) would have resulted in rapid cooling to subsolidus temperatures where autometasomatic reactions involving evolved deuteric fluids would dominate as indicated by the chloritization of biotite. We suggest that these subsolidus reactions were unlikely to have involved meteoric fluids as (i) all of the olivine macrocrysts and phenocrysts in these rock types are replaced by calcite and not serpentine which cannot be explained by hydrothermal alteration involving hydrous meteoric fluids, and (ii) all of the primary groundmass perovskite is significantly altered to Mn-ilmenite + titanite + pyrite (figure 3.5E) which is characteristic of alteration by CO2-rich fluids as will be discussed in much greater detail in Sect. 5.9 Perovskite Alteration by CO2-rich Fluids. It can therefore be concluded that the KPK and KPKt rock types in Kimb65a represent the solidified products of a relatively anhydrous kimberlite melt and do not display textures or mineralogy that would support the interpretation of post-emplacement hydrothermal alteration. 5.6.2 Reactions between Crustal Xenoliths in HK, HKt, and KPKt Rock Types in Kimb65b and Kimb65d The crustal xenoliths in HK, HKt, and KPKt rock types in Kimb65b and Kimb65d contain (i) a fine-grained assemblage of pectolite + serpentine + calcite, replacing the primary assemblage of quartz + feldspar + biotite, and (ii) thin (0.5 – 1 mm) coronas on large (> 2 cm) xenoliths or affecting the entire clast for xenoliths < 2 cm (figure 3.6, 3.15 & 5.4); the coronas contain an inner rim of diopside and an outer rim of phlogopite. The host kimberlite in these transitional rock types appears mottled pale reddish-brown in colour in macrospecimen (figure 5.4) in haloes surrounding the reacted crustal xenoliths. These hybrid patches are composed of diopside + phlogopite + serpentine in contrast to the assemblage of phlogopite  Discussion - Crustal Xenoliths in Renard 65  98   Figure 5.4 – Polished slab photograph of an HKt rock type from Kimb65b illustrating the reaction of crustal xenoliths, the mottled light reddish brown patches of diopside and phlogopite rich hybrid kimberlite groundmass, and the relationship between olivine macrocryst serpentinization and proximity to crustal xenoliths. + serpentine + calcite further away from crustal xenoliths in the darker non-hybrid groundmass (figure 5.4). The serpentinization of olivine macrocrysts and microcrysts appears to be correlated with the development of these haloes around crustal xenoliths as olivine further from the xenoliths in the more typical groundmass assemblage are generally fresh to partially serpentinized while those closer to crustal xenoliths are fully serpentinized (figure 3.6A & 5.4). Described below are the reactions which are believed to have taken place at suprasolidus temperatures producing the recrystallized margins of the xenoliths and the hybrid groundmass, followed by metasomatic reactions which are believed to have taken place at subsolidus temperatures resulting in the crystallization of monomineralic coronas. Discussed separately is the development of aegirine zoning on the diopside coronas which is related to both the serpentinization of olivine and the compositional modification of groundmass spinels.  The initial reaction between entrained crustal xenoliths and the host kimberlite is described by reaction (13) in which quartz, orthoclase, albite and biotite recrystallize to an assemblage of pectolite, serpentine and calcite, by addition of the ubiquitous components CO2, H2O, MgO and CaO from the kimberlite melt (figure 5.5A & 5.5B). The subscript (m) in reactions 13 and 14 denotes components in melt. Discussion - Crustal Xenoliths in Renard 65  99   Figure 5.5 – (A) The initial reaction between fresh granitoid xenoliths and the host kimberlite magma is characterized by the addition of CO2, H2O, MgO and CaO and (B) removal of SiO2, Al2O3, FeO and K2O to produce the reaction mineral assemblage pectolite + serpentine + calcite. The contamination of the kimberlite magma by this initial reaction results in a shift in stable groundmass mineralogy in a halo around the crustal xenoliths from phlogopite + serpentine + calcite to diopside + phlogopite + serpentine. Abbreviations are given as quartz (qtz), orthoclase (or), albite (ab), biotite (bt), phlogopite (phl), serpentine (srp), calcite (cal), pectolite (pct). Discussion - Crustal Xenoliths in Renard 65  100  qtz   +   or   +   al   +   bt   +   CO2/H2O/MgO/CaO   →   pct   +   srp   +   cal   +   SiO2/Al2O3/FeO/K2O SiO2 + KAlSi3O8 + NaAlSi3O8 + K(Mg,Fe)3AlSi3O10(OH)2 + 3CO2(m) + 2H2O(m) + 2MgO(m) + 7CaO(m) →              2NaCa2Si3O8(OH) + Mg3Si2O5(OH)4 + 3CaCO3 + 5SiO2(m) + 2Al2O3(m) + 2FeO(m) + K2O(m) The addition of the components SiO2, Al2O3, FeO and K2O released by reaction (13) into the adjacent kimberlite melt shift the stable groundmass assemblage from phlogopite + serpentine + calcite to the hybrid assemblage diopside + phlogopite + serpentine as described by reaction (14) (figure 5.5B & 5.5C). phl   +   cal   +   srp   +   SiO2   +   MgO   →   di   +   srp   +   phl   +   CO2 KMg3AlSi3O10(OH)2   +   CaCO3   +   Mg3Si2O5(OH)4   +   2SiO2(m)   +   MgO(m) →                           (14) CaMgSi2O6   +   Mg3Si2O5(OH)4   +   KMg3AlSi3O10(OH)2   +   2CO2 These reactions are proposed to have occurred at high temperature by reaction between cold crustal xenoliths at ambient temperature (< 100 °C) and hot kimberlite melt (< 900 °C) on the basis that the hybrid groundmass contains significantly lower (approximately 50% lower) modal abundances of primary spinel and perovskite (figure 3.10E, 3.10F, 3.17E & 3.17F). If the hybrid groundmass assemblage was the result of subsolidus processes (i.e. hydrothermal alteration), the modal abundances of these early-formed groundmass minerals ought to be consistent in both the hybrid and non-hybrid groundmass assemblages. A reduction in these modal abundances can be attributed to increased crystallization of diopside and phlogopite from the hybrid melt in response to the addition of the components SiO2, Al2O3, FeO and K2O released by reaction (13). The presence of serpentine within the xenolith reaction rims does not preclude high temperature reaction between xenoliths and kimberlite magma despite the fact that serpentine is unstable at temperatures above 600 °C (O’Hanley, 1996), which is below the kimberlite solidus. The heat transferred from the hot kimberlite magma to the cold crustal xenolith would result in a thermal gradient and require partial crystallization of the magma around the xenolith (Winter, 2010). The temperatures within the recrystallized margins of the crustal xenoliths would have been somewhere between the Discussion - Crustal Xenoliths in Renard 65  101  ambient temperature of the country rock and the temperature of the kimberlite melt, varying by the size and composition of the xenolith. As the kimberlite cooled to solidus temperatures, subsolidus metasomatism involving deuteric fluids resulted in the formation of monomineralic rims of diopside (variably zoned to aegirine) and phlogopite (figure 3.6, 3.7 & 3.15).  The formation of monomineralic zones at the interfaces between bulk compositions in disequilibrium is the hallmark feature of metasomatism; the zones crystallize in response to chemical potential gradients either without the assistance of a fluid phase, or with diffusion enhanced by either a stationary or infiltrating fluid phase (Winter, 2010). We propose that the formation of monomineralic rims of diopside and phlogopite are likely to represent subsolidus metasomatism in response to chemical potential gradients in Mg and Si between the silicic recrystallized margins of the crustal xenoliths and the Mg-rich hybrid kimberlite groundmass (figure 5.6A & 5.6B); the occurrence of these zones at the interface between the recrystallized margins of the xenoliths and the hybrid kimberlite groundmass suggests that they grew at the expense of either or both of those zones (figure 5.6B). Since the chemical potential gradients on either side of the contact are fixed, the growth of the monomineralic rims cause a reduction in the chemical potential gradient until a state of local equilibrium is established (figure 5.6B) (Winter, 2010). For simplicity the changes in bulk composition and chemical potential gradients are illustrated only for Si (figure 5.6), but similar and inverse trends could be illustrated for Mg. To show that the bulk compositional changes across of each of these zones (figure 5.6B) do indeed demonstrate metasomatism in response to these chemical potential gradients, each of the zones including the monomineralic rims are illustrated in a chemographic Na, K, Ca – Mg, Fe – Si diagram (figure 5.7). Each zone is represented on the chemographic diagram as a triangle containing all of the minerals within the assemblage of that zone: zone 1 refers to the non-hybrid kimberlite groundmass, zone 2 refers Discussion - Crustal Xenoliths in Renard 65  102   Figure 5.6 – (A) The recrystallized crustal xenoliths and the hybrid kimberlite melt define two compositionally intermediate zones between the unreacted crustal xenolith assemblages and the unmodified kimberlite groundmass assemblages. The bulk composition of each zone records a steady increase in wt% SiO2 and a steady decrease in MgO from the xenolith into the kimberlite, with disequilibrium at the contacts between zones resulting in steep chemical potential gradients. (B) In the subsolidus environment, bimetasomatic reactions are likely enhanced by intergranular fluids resulting in the diffusion of SiO2 out of the xenoliths and MgO into the xenoliths. This diffusion results in the growth of monomineralic reaction rims, which reduce the steepness of the chemical potential gradient with increased growth until a state of local equilibrium is achieved. to the hybrid kimberlite groundmass, zone 3 refers to the phlogopite monomineralic rim, zone 4 refers to the diopside (± aegirine) monomineralic rim, zone 5 refers to the recrystallized margins of the xenoliths and zone 6 refers to the unreacted cores of the xenoliths as were depicted in figure 5.6. The chemographic diagram clearly illustrates that the changes in bulk composition across the entire sequence generally reflect correlate with diffusion of the components SiO2 and MgO. Discussion - Crustal Xenoliths in Renard 65  103   Figure 5.7 – Chemographic Na, K, Ca – Mg, Fe – Si diagram illustrating changes in bulk composition from the uncontaminated kimberlite groundmass (1) to the unreacted cores of crustal xenoliths (6). 1: uncontaminated groundmass (ol + srp + phl + sp + cc). 2: hybrid kimberlite groundmass (ol + srp + phl + di + sp). 3: monomineralic phlogopite corona. 4: monomineralic diopside/aegirine corona. 5: recrystallized crustal xenolith (pct + cc + srp). 6: unreacted crustal xenolith (or + alb + bt + qtz). Dashed lines indicate mineral parageneses for each mineralogical zone, and solid lines connect the bulk compositions of each zone.  One final feature of these metasomatic rims that warrants further discussion is the zoning of fine-grained diopside in the monomineralic coronas to much coarser aegirine protruding into the cores of the xenoliths (figure 5.8). This zoning is visible in xenoliths of all rock types in Kimb65b and Kimb65d, but becomes increasingly well developed from HKt rock types in Kimb65b (figure 3.7E & 3.7F) to KPKt rock types in Kimb65d (figure 3.15D). The extent to which aegirine zoning is developed is correlated with both increasing degrees of serpentinization of olivine (figure 3.6A & 5.4) and also with changes in groundmass spinel composition from chromite to magnesiochromite (figure 3.21).  We propose that diopside-aegirine zoning records an increase in the availability of Fe3+ during the late stages of metasomatism which must have been more locally available in transitional rock types or Discussion - Crustal Xenoliths in Renard 65  104   Figure 5.8 – (A) PPL and (B) XPL thin-section micrographs illustrating the compositional zoning from diopside to aegirine protruding into the core of the xenolith from the otherwise monomineralic diopside corona on the outer margin of the xenolith. The zoning from diopside to aegirine is associated with a significant increase in the crystal sizes suggesting late-stage slow growth. transitional zones in coherent rock types. The availability of Fe3+ is likely directly related to the localized effects of olivine serpentinization, as reactions involving the hydration of olivine typically involve oxidation of Fe2+ to Fe3+ and the production of H2 (Evans et al., 2004); in serpentinites the available Fe3+ is commonly incorporated into magnetite (Fe2O3) and to a lesser extent into serpentine. In some kimberlites, magnetite crystallizes around serpentinized olivines texturally resembling necklace crystallization of spinel, but in the HK, HKt and KPKt rock types in Kimb65b and Kimb65d magnetite does not crystalize around serpentinized olivine. It is suggested that Fe3+ was primarily consumed in the late-crystallization of aegirine zoned on diopside (figure 5.8). It is unlikely that the activity of Na would have been the driving factor in the crystallization of aegirine. Na is not a characteristic component of kimberlite melts but is derived in the Renard 65 kimberlites from crustal xenoliths; so the availability of Na ought to be higher in magmas that solidify to produce coherent rock types where higher temperatures and prolonged durations of interaction should lead to greater degrees of assimilation.  While the increased speciation of Fe3+ was likely a consequence of the localized effects of olivine serpentinization and was the primary cause of the aegirine zoning, the overall increase in the activity of Fe in the fluid phase is attributed to the redistribution of Fe from groundmass spinels into clinopyroxene Discussion - Crustal Xenoliths in Renard 65  105  in the reaction rims of the crustal xenoliths. In Sect. 3.2.4 Spinel Mineral Chemistry, it was shown that the Fe #’s in groundmass spinels decreased systematically from high Fe #’s in HK rock types in Kimb65b, to low Fe #’s in KPKt rock types in Kimb65d, with a complete range of intermediate Fe #’s in HKt rock types in Kimb65b (figure 3.21). It is therefore possible to explain the increased activity of Fe as the result of increased subsolidus exchange of Mg for Fe in spinels hosted within increasingly transitional rock types. We suggest that more transitional rock types would exsolve more fluids during the final stages of cooling than coherent rock types as a result of Si contamination and hybrid groundmass crystallization. The increased abundance of an exsolved volatile fluid phase would provide conditions for subsolidus spinel alteration by a fluid with which it is in disequilibrium, releasing Fe which combines with available Na to stabilize the crystallization of aegirine in the metasomatic coronas on crustal xenoliths as described by reaction (15). di + sp + ol + H2O(aq) + Na+(aq) + SiO2(aq) → aeg + sp + serp + Ca2+(aq) + H2(g) CaMgSi2O6 + Fe2+Cr2O4 + 2(Mg,Fe)2SiO4 + 3H2O(aq) + 2Na+(aq) + 2SiO2 →       (15) 2NaFe3+Si2O6 + MgCr2O4 + (Mg,Fe)3Si2O5(OH)4 + Ca2+(aq) + H2(g) These subsolidus reactions involving the formation of monomineralic metasomatic coronas, the serpentinization of olivine, and alteration of groundmass spinel chemistry are best characterized by interaction with deuteric fluids rich in Mg. Alteration is unlikely to have involved any significant influx of post-magmatic meteoric fluids for the following reasons: 1) Aegirine is commonly observed in sodic fenites surrounding carbonatite intrusions (Le Bas, 2008). Sodic fenites are metasomatic aureoles resulting from the release of alkali-rich fluids during the late-stages of crystallization of sovite intrusions. The presence of aegirine zoning on diopside in monomineralic rims of the Renard 65 kimberlites therefore indicates that not only are the coronas metasomatic, but that they are the result of deuteric, CO2-rich fluid interaction. Discussion - Crustal Xenoliths in Renard 65  106  2) The availability of Fe3+ in circulating meteoric fluids should not vary depending on the rock type/texture. If metasomatism in the Renard 65 kimberlites was the result of hydrothermal alteration by meteoric fluids, aegirine-to-diopside zoning would be expected to form equally within diopside coronas in xenoliths throughout each of the HK, HKt and KPKt rock types. Furthermore, it should also affect KPK/KPKt rock types in Kimb65a which it does not. 3) The fluid compatible with diopside should not necessarily be CO2-poor. The monomineralic coronas on crustal xenoliths in Renard 65 are not constrained to very low XCO2 fluid compositions as were hydrogarnet- and bultfonteinite-bearing assemblages in metasomatic coronas on crustal xenoliths in volcaniclastic kimberlites from Orapa (Buse et al., 2010); diopside can crystallize from a Group I kimberlite melt at temperatures below 1200 °C and XCO2 = 0.24 at 10 kbar (Edgar et al., 1998). Both diopside and phlogopite almost certainly crystallized at subsolidus temperatures but neither preclude metasomatism by a CO2-bearing deuteric fluid. 4) Deuteric fluids would be readily available and inherently Mg-rich before and particularly after serpentinization. 5) The country rock at Renard does not include any permeable sedimentary lithologies that could provide a source for meteoric fluid influx and circulation as described at Venetia and Orapa (Brown et al., 2008; Field et al., 2008). It is reasonable to conclude that the reactions between crustal xenoliths and their host kimberlites in Renard 65 are characterized by both in-situ magmatic reactions and subsolidus autometasomatic reactions. The features of both styles of reaction preclude hydrothermal alteration by circulating meteoric fluids. While hydrothermal alteration may provide a viable explanation for the observed features of some kimberlite pipes where the geological features of the kimberlite and host country rock support such an interpretation, it is suggested that these features and the implied process of formation are not characteristic of a large number of KPK pipes.  Discussion - Degree of Assimilation of Crustal Xenoliths in Kimberlites  107  Since the reactions between crustal xenoliths and their host kimberlite in the Renard 65 kimberlites are arguably at least initiated in a magmatic environment, there exists the potential for crustal xenolith assimilation by kimberlite melts. In the following sections we will model the potential for contamination of kimberlite melt by silicic xenoliths and examine the possible effects that contamination may have on the exsolution of magmatic volatiles during emplacement and the formation of transitional and KPK rock textures. 5.7 Degree of Assimilation of Crustal Xenoliths in Kimberlites From a general volcanology and igneous petrogenesis point of view, most mantle derived magmas upon passing through the crust during their ascent will react with and assimilate wall-rock, and entrain, partially melt, and assimilate crustal xenoliths (Parfitt, 2008). Assimilation is most predominant within magma chambers as a result of latent heat production from fractional crystallization and extensive residence times, and thus crustal magmatism is often modelled as assimilation-fractional-crystallization (AFC) (Bowen, 1928; Wilcox, 1954; DePaolo, 1981). Since assimilation is an endothermic process requiring heat transfer from the magma to the xenoliths and wall rock, the typical range of degrees of assimilation of 5-21 % for basaltic magmas (with a liquidus temperature of 1230°C) are a function of thermal constraints including: the heat flux across the xenolith and the energy required for melting the xenolith (McLeod & Sparks, 1998). The heat flux to the xenolith depends on the thermal gradient and conductivity of the melt layer interface between the xenolith and the host melt and the temperature of the host melt, while the energy required the melt the xenolith is related to the density, specific heat, melting temperature, enthalpy of melting, and the initial temperature of the xenolith (McLeod & Sparks, 1998).  Using these general thermal constraints, and assuming crustal xenolith assimilation in basaltic magmatism to be analogous to that in kimberlite magmatism, it is possible to suggest an appropriate range of values for the degree of assimilation of crustal xenoliths in kimberlite magmas. Kimberlite Discussion - Degree of Assimilation of Crustal Xenoliths in Kimberlites  108  magmas erupt from narrow feeder dykes experiencing turbulent flow, typically 0.3 – 1 m in width and 0.5 – 10 km in length, comparable to dyke dimensions in basaltic magmatism (Sparks et al., 2006), and eruption temperatures between 1030 - 1170°C (Fedortchouk & Canil, 2004; Phillips & Harris, 2003; Sparks et al., 2006, 2009; Sparks, 2013; Kavanagh & Sparks, 2009), comparable to the 1230°C liquidus temperature of MORB (Sparks et al., 2006). Unlike basaltic magmas however, kimberlite magmas are very low viscosity (0.1 – 1 Pa s) with very fast ascent rates of 4 – 20 m/s and typical magma supply rates of 102 – 105 m/s3 generating Plinian intensity volcanism with a duration of hours to months (Sparks et al., 2006). In addition, kimberlite magmas being silica-undersaturated, are depolymerized and have very fast crystallization kinetics (Kavanagh & Sparks, 2009); consequently, at shallow (5 - 10 km) depths, the cooling effects of volatile degassing, which is an endothermic process (Russell, 1987) are buffered by the release of latent heat from rapid crystallization resulting in a net rise in temperature (Kavanagh & Sparks, 2009). Based on the similarities in dyke dimensions and emplacement temperatures, crustal xenolith assimilation rates may be generally comparable between kimberlite and basalt magmas. While kimberlite magmas are not believed to form magma chambers, and consequently do not have the added contribution of latent heat from fractional crystallization and prolonged residence times, kimberlite magmas do have significant latent heat contribution from crystallization as a consequence of volatile degassing during emplacement through the upper crust (5 – 10 km depth). It is here assumed that these thermal constraints act to counter one another, and an upper limit of 20% for crustal xenolith assimilation is appropriate for kimberlite magmas; a similar minimum value of 8 - 12 % was suggested by Caro et al. (2004) based on hybrid Sr - Nd isotopic signatures from the assimilation of crustal xenoliths by primary kimberlite melt. It is suggested that this upper limit of 20% should only apply to HK rock types in dykes or in the root zones of KPK pipes, with lower degrees of assimilation expected for transitional to upper diatreme facies HKt, KPKt and KPK rock types. The source temperature of pyroclastic kimberlite deposits are estimated to be approximately 760 - 920°C as a result of cooling from magmatic temperatures of 1030 – 1170 °C after the introduction Discussion - Effects of Crustal Xenolith Assimilation on Volatile Fluid Exsolution in Kimberlite Melt  109  of ~ 10 – 20 % cold lithic clasts (Fontana et al., 2011; Pell et al., 2015); these values are consistent with maximum temperatures of  800 - 1000 °C suggested by the amphibole reaction mineral assemblages, down to minimum temperatures of 180 – 350 °C suggested by the chlorite + pyrite reaction mineral assemblages in xenoliths in KPK/KPKt rock types in Renard 65 (Sect 5.4 Crustal Xenoliths in Renard 65). It can therefore be suggested that the degree of assimilation for crustal xenoliths in various textural facies of KPK pipes may range from an upper limit of 20% assimilation in HK dykes and root zone HK rock types, approaching as little as 1% for the upper diatreme KPK rock types.  5.8 Effects of Crustal Xenolith Assimilation on Volatile Fluid Exsolution in Kimberlite Melt To model the potential effects of xenolith assimilation on kimberlite melt composition and consequently volatile exsolution, two primary kimberlite melt compositions at 20.77 wt % SiO2 (JD 51) and 31.86 wt % SiO2 (LGS07-3) are modelled at 5, 10, and 20 % assimilation for 20, 40, 60 and 80 % volumetric proportions (modal abundances) of crustal xenoliths (Table 5.1). Kimberlite melt compositions are from aphanitic (JD 51) and macrocrystic (LGS07-3) kimberlites from Jericho (Price et al., 2000) as primary melt compositions for the Renard 65 kimberlites are not available, and their determination is beyond the scope of this study. The modelled xenolith composition (CR) is a 1:1 average whole rock composition for granitic and gneissic in-situ country rock at Renard (unpublished data courtesy of Stornoway Diamond Corporation). The selection of the two kimberlite melt compositions represents the high and low SiO2 end-members of published and publically available primary kimberlite melt compositions. Both melt compositions represent kimberlites that contain negligible quantities of crustal xenoliths and thus are very unlikely to have experienced contamination. It has also been proposed that kimberlites may partially assimilate mantle xenocrysts (orthopyroxene in particular), resulting in increased a(SiO2) and reducing the solubility of CO2 (Russell et al., 2013). The composition of the macrocrystic kimberlite end-member (LGS07-3) which contains a typical modal abundance of olivine macrocrysts (mantle derived olivine xenocrysts), must account for any possible mantle xenolith or xenocryst assimilation. Therefore, the two Discussion - Effects of Crustal Xenolith Assimilation on Volatile Fluid Exsolution in Kimberlite Melt  110  kimberlite melt compositions used in the model are representative of the expected minimum and maximum SiO2 end-members (and associated maximum CO2 contents as a function of SiO2 content) of kimberlite magmas at the time of emplacement. These modeled hybrid melt compositions (Table 5.1) show that for Si-poor kimberlite melts (~20 wt % SiO2), contamination by crustal xenoliths could result in as little as a 1 % increase in wt % SiO2 for a 20 % crustal xenolith modal abundance and a 5 % degree of assimilation, up to as much as a 35 % increase in wt % SiO2 for an 80 % crustal xenolith modal abundance and a 20 % degree of assimilation. For an Si-rich kimberlite melt composition (~32 wt% SiO2), contamination by crustal xenoliths could result in as little as a 2 % increase in wt % SiO2 for a 20 % crustal xenolith modal abundance and a 10 % degree of assimilation, up to as much as a 17 % increase in wt % SiO2 for an 80 % crustal xenolith modal abundance and a 20 % degree of assimilation. In addition to greatly increased a(SiO2), hybrid melt compositions will likely have increased activities of Al, Fe, Na and K (Table 5.1). These increases in the SiO2 content of the kimberlite melt have direct consequences on the exsolution of volatile CO2 and H2O fluids during emplacement in the upper crust (< 10 km depth). It is well established that the (i) solubility of CO2 in silicate and transitional (intermediate between silicate and carbonatitic) melts decreases in response to increasing wt% SiO2 in the melt, and (ii) that the solubility of CO2 and H2O decrease as a function of decreasing pressure (Blank & Brooker, 1994; Brey & Ryabchikov, 1994; Brooker et al., 2001; Russell et al., 2012). Moussallam et al. (2015) analyzed quenched pure glasses of synthetic transitional melts with 25 – 32 wt% SiO2 and 53 – 63 wt% MgO + CaO (calculated on a volatile free basis) in a depolymerized configuration and showed that the solubility of CO2 in kimberlite melts increases nearly exponentially with increasing pressure, in contrast to the linear relationship between CO2 solubility and pressure in basaltic magmas. Moussallam et al. (2016) showed further that the solubility of H2O in transitional melts is similar to the solubility of H2O in basaltic melts, displaying a power law relationship with ƒ(H2O); their results show that small changes in Si or H2O content will affect the depth at which intensive volatile exsolution occurs. Discussion - Effects of Crustal Xenolith Assimilation on Volatile Fluid Exsolution in Kimberlite Melt  111  Table 5.1 – Modelled hybrid kimberlite melt compositions for minimum and maximum initial SiO2 kimberlite melt compositions resulting from various degrees of assimilation of different modal abundances of crustal xenoliths.   Discussion - Effects of Crustal Xenolith Assimilation on Volatile Fluid Exsolution in Kimberlite Melt  112   Figure 5.9 - H2O-CO2 dependency experimentally determined by Moussallam et al. (2016) compared with the model of Papale et al. (2006).  The H2O-CO2 dependency that they determined (figure 5.9) suggests that melts with higher XH2O will exsolve a greater proportion of CO2 at a greater depth compared to melt with lower XH2O.. For the kimberlite rock types that comprise the main geological units in Renard 65, the modal abundances of crustal xenoliths and estimated degrees of assimilation are used to estimate the potential for contamination (Table 5.2). These calculations assume that (i) each geological unit represents a kimberlite magma, and (ii) that the initial compositions of each magma are the same (either Si-rich or Si-poor initial kimberlite melt compositions). It is in fact very unlikely that each of these magmas would have  Table 5.2 – Calculated minimum and maximum increases in wt% SiO2 for the Renard 65 geological units  Discussion - Effects of Crustal Xenolith Assimilation on Volatile Fluid Exsolution in Kimberlite Melt  113  the same initial composition, but in determining the potential for contamination, exact initial compositions are irrelevant. These calculations indicate that the KPKt rock types in Kimb65d should show the highest degree of contamination, with increases of 6 – 13 wt% SiO2, followed by HK/HKt rock types in Kimb65d with increases of 4 – 9 wt% SiO2. Much lower degrees of contamination are predicted for KPK rock types of Kimb65a with increases of 3 – 5 wt% SiO2, and for HK rock types of Kimb65c with increases of 1 – 2 wt% SiO2. Based on established relationships between CO2 solubility as a function of pressure for melts with various wt% SiO2 content (Moussallam et al., 2015), these hypothetical relative increases in wt% SiO2 in the Renard 65 kimberlites translate into potential exsolution of 0.8 wt% CO2 for Kimb65a, 1.4 wt% CO2 for Kimb65b, 0.4 wt% CO2 for Kimb65c, and 2.0 wt% CO2 for Kimb65d (figure 5.10) at a pressure equivalent   Figure 5.10 – Assimilation of crustal xenoliths and hybridization of kimberlite melts to higher Si content reduces the solubility of CO2 and triggers exsolution. Each of the Renard 65 geological units are modelled for their relative potential to exsolve CO2 in response to crustal xenolith assimilation. The median increases in wt % SiO2 for each geological unit (Table 5.2), are transposed onto experimentally derived curves from Moussallam et al. (2015), showing the solubility of CO2 as a function of pressure for synthetic melts with various Si contents. Greater relative exsolution of CO2 is expected for the kimberlite magma(s) that formed kimberlites within the geological units Kimb65b and Kimb65d (11 – 16 % of the total remaining dissolved CO2 at a pressure equivalent depth of 3 km). Much lower relative exsolution of CO2 is expected for the kimberlite magma(s) that formed kimberlites within the geological units Kimb65a and Kimb65c (3 – 7 % of the total remaining dissolved CO2 at a pressure equivalent depth of 3 km). Discussion - Perovskite Alteration by CO2-rich Fluids  114  depth of 3 km. Therefore, a significant amount (4 – 20 %) of the CO2 exsolved during the final 3 km of emplacement may be in response to the assimilation of crustal xenoliths, and not entirely in response to decompression. This assimilation and CO2 solubility model suggests that in as much as magmatic volatile degassing, authigenic brecciation and fluidization are viable processes for the formation of the textures and features associated with KPK rock types, the assimilation of crustal xenoliths may be nearly as important in producing those features as a result of magmatic volatile exsolution during decompression. The importance of these components in driving the stability of common interclast matrix assemblages was described by reaction (1) in Sect. 5.3 Current Models on the the Origin of the Interclast Matrix in KPKs. In fact, it could be argued that the contamination of kimberlite melt by felsic material is a necessary step in the formation of KPKs because (i) assemblages of diopside + phlogopite + serpentine are not observed within FPK rock types that have also undergone degassing and fragmentation (Scott Smith, 2008), and (ii) phlogopite is absent the interclast matrix of pyroclastic deposits at Igwisi Hills which are interpreted to represent modern kimberlites, and these are not contaminated by felsic xenoliths (Willcox et al., 2015). We further highlight that this conclusion is consistent with the features observed in other KPK localities beyond Renard such as Gahcho Kue. Hetman et al. (2004) note that the distribution of xenoliths in the Gahcho Kue pipes decreases in modal abundance with depth, simultaneous with transitions from KPK to KPKt to HKt to HK rock types. They attribute this xenolith distribution to the buoyancy effects of upwelling exsolved volatiles. We suggest instead that the correlation between high felsic xenolith abundances and KPK textures is evidence of the importance of xenolith assimilation in the formation of KPK textures.  5.9 Perovskite Alteration by CO2-rich Fluids In Sect. 5.6.1 Crustal Xenoliths in Renard 65 – Reactions Between Crustal Xenoliths in KPK Rock Types in Kimb65a, it was shown that the reactions between crustal xenoliths and KPK and KPKt rock types Discussion - Perovskite Alteration by CO2-rich Fluids  115  in Kimb65a occurred over a wide range of temperatures (180 to < 800 °C) involving relatively low XH2O fluids.  This interpretation was supported by (i) relatively low abundances of groundmass phlogopite (ii) the pseudomorphing of olivine by calcite throughout the rocks, and (iii) the alteration of primary perovskite in magmaclasts to Mn-ilmenite + titanite + pyrite.  Here we discuss in more detail the how the alteration of perovskite helps to constrain the CO2 rich nature of fluids interacting with the kimberlite at subsolidus temperatures. The suggestion that the kimberlite melt which produced KPK and KPKt rock types in Kimb65ba was relatively anhydrous will become important Sect. 5.11 Geological Modelling of the Renard 65 pipe; specifically, the development of one of the more unusual features of Kimb65a in Renard 65 (that has also been observed in the Renard 2 pipe), the immense vertical extent of KPK and KPKt textures to significant depth within the pipe and the seemingly absent root zones. The primary mode of occurrence of perovskite (CaTiO3) in kimberlite is as a euhedral to subhedral crystal, typically ranging in size from 10 to 50 µm and evenly distributed throughout a late mesostasis groundmass on the order of < 10 % modal abundance (Mitchell, 1986). Crystals may exhibit compositional zonation with cores enriched in LREE, Th and Nb, decreasing towards the rims (Chakhmouradian & Mitchell, 2000); in some kimberlites this zonation pattern is reversed, or crystals show oscillatory zonation (i.e. at the Benfontein Sill in South Arica or the Jericho Pipe in Canada) (Chakhmouradian & Mitchell, 2000; Kopylova & Hayman, 2008). Perovskite becomes unstable at temperatures below 350 °C and pressures below 2 kbar, in the presence of CO2-rich fluids (Mitchell & Chakhmouradian, 1998). The most common replacement products are the TiO2 polymorphs anatase and rutile with calcite, via a Ca-leaching replacement reaction (16) (Nesbitt et al., 1981).  CaTiO3 + CO2 ⇌ TiO2 + CaCO3      (16) In the Chromur and Iron Mountain kimberlites in Yakutia and Wyoming, anatase and rutile replacing perovskite are intergrown with titanite (CaTiSiO5) (Mitchell & Chakhmouradian, 2000); the crystallization Discussion - Perovskite Alteration by CO2-rich Fluids  116  of titanite in kimberlite groundmass is uncommon (Mitchell, 1986) and inconsistent with the very Si-poor compositions of primary kimberlite melts (18-32 wt% SiO2) (Price et al., 2000; Kopylova et al., 2007; Moussallam et al., 2016). The intergrowths of titanite replacing perovskite are attributed to a Ca-leaching reaction (17) by a CO2-rich fluid with a high a(SiO2) (Mitchell & Chakhmouradian, 2000). Mitchell and Chakhmouradian (2000) suggest that anomalously high a(SiO2) would most likely be attributable to assimilation of crustal xenoliths, but did not offer petrographic evidence in support of this theory. 2CaTiO3 + CO2 + (SiO2) (liquid) ⇌ TiO2 + CaTiO5 + CaCO3    (17) After anatase and rutile, the second most common replacement of perovskite is by ilmenite. In most of these occurrences (e.g. Chromur in Yakutia), ilmenite appears to be a successive replacement after anatase and rutile, but in some deposits (e.g. Tunraq and Chicken Park kimberlites in Nunavut and Wyoming) the replacement of perovskite by ilmenite lacks intermediate TiO2. Ilmenite replacing perovskite in these occurrences shows significant enrichment of Mn (up to 31 mol % MnTiO3), depletion of Mg (< 2 mol % MgTiO3) and very low Cr or Fe3+ (Mitchell & Chakhmouradian, 2000). The direct replacement of perovskite by Mn-enriched ilmenite involves Ca-leaching in the presence of high a(Mn2+) or a (Fe2+) (reaction 18) (Mitchell & Chakhmouradian, 2000): CaTiO3 + (1-n)Fe2+ + nMn2+ ⇌ Fe1-nMnnTiO3 + Ca2+    (18) The preference for Fe (Fe2+TiO3) over Mg (MgTiO3) ilmenite end-members replacing perovskite (as observed in Tunraq, Chicken Park) are a consequence of the relative activities of Fe2+ and Mg2+ in the deuteric fluid during replacement; the equilibria between perovskite and the two ilmenite group minerals illustrated by isobaric curves (figure 5.11), show that the replacement of perovskite by Fe2+TiO3 requires a relatively low a(Fe2+) [log a(Ca2+)/a(Fe2+) > 6-10] whereas the replacement of perovskite by MgTiO3 requires a significantly higher a(Mg2+) [log a(Ca2+)/a(Mg2+) > 1] (Chakhmouradian & Mitchell, 2000). In addition, the minimum required a(Fe2+) in the fluid to crystallize Fe2+TiO3 decreases rapidly with  Discussion - Perovskite Alteration by CO2-rich Fluids  117   decreasing temperature, while the replacement of perovskite by MgTiO3 is essentially temperature independent. The final and least common replacement product of perovskite is pyrite, present as an intergrowth in multicomponent pseudomorphs, as documented in sulfide-rich kimberlites such as Peuyuk C kimberlite (Nunavut). Pyrite intergrowths replacing perovskite are attributed to high a(Fe2+) and a(H2S) in the deuteric fluids (Mitchell & Chakhmouradian, 2000).  In the Renard 65 kimberlites, perovskites display a range of zonation and replacement patterns. These patterns are consistent throughout each geological unit. Perovskite in magmaclasts in KPK and KPKt rock types from Kimb65a are either partially or fully replaced by Mn-enriched ilmenite (Fe2+TiO3) along the outer margins of the crystal, with calcite crystallizing in the void-space between the Mn-ilmenite rims and the partially preserved perovskite cores (figure 3.5E). In almost all of the observed magmaclasts, perovskite replacement occurs throughout the entire magmaclast; however, in some very rare occurrences zoned magmaclasts are observed in which perovskite preservation is higher in the core and Figure 5.11 – Perovskite equilibria with ilmenite-group minerals illustrated by isobaric curves at P = 1 MPa (after Chakhmouradian & Mitchell, 2000) Discussion - Perovskite Alteration by CO2-rich Fluids  118  lower along the margins (figure 3.2A & 3.2B). These observations suggest that the replacement of perovskite occurred syn- or post-magmaclast formation, affecting primary magmaclast groundmass assemblages zonally from the margin of the clast towards the core. In addition, these magmaclast groundmass assemblages also contain anhedral interstitial titanites, and subhedral to anhedral pyrite (figure 3.5C, 3.5D & 3.5F). Although pyrite and titanite do not occur as intergrowths with ilmenite directly replacing perovskite, their co-occurrence in the groundmass of the magmaclasts suggests they are likely to be genetically related to the replacement of perovskite by Mn-ilmenite. Perovskite in the interstitial groundmass of HK rock types from Kimb65c are replaced in a manner very similar to those in Kimb65a, with Mn-ilmenite (Fe2+TiO3) forming mantles on remnant perovskite cores (figure 3.14C). However, in HK rock types of Kimb65c, pyrite and titanite are observed both as intergrowths with Mn-ilmenite replacing perovskite (figure 3.14D & 3.14E) and also as acicular or anhedral and interstitial crystals within the groundmass. These observations provide very strong evidence that the occurrence of Mn-ilmenite intergrown or co-occurring with pyrite and titanite are all related to the replacement of perovskite. By comparison, perovskite in HK and HKt rock types in Kimb65b and KPKt rock types in Kimb65d remains pristinely preserved. It can therefore be concluded that subsolidus fluids interacting with KPK and KPKt rock types during or after the formation of magmaclasts were of a relatively anhydrous character; this may be the result of either or both (i) an initially high CO2/H2O ratio in the kimberlite melt, or (ii) very low degrees of crustal xenolith contamination (Table 5.2) allowing for greater CO2 solubility to shallow depths. It is argued that while both are likely to be true, the low abundance of phlogopite in magmaclasts, the alteration of groundmass perovskite, and the pervasive replacement of olivine by calcite all indicate that the melt was likely characterized by an initially high CO2/H2O ratio.  Discussion - Oxygen Fugacity  119  5.10 Oxygen Fugacity This discussion has focused primarily on the viability of different emplacement processes in explaining the features of KPK rock types at Renard 65. Before summarizing these features and proposing an emplacement process (including the identification of separate magmatic events or phases of kimberlite) we will constrain one final intrinsic parameter of kimberlite magmas just prior to emplacement, their oxygen fugacity. Oxygen fugacity is a thermodynamics parameter that describes the redox state of a medium relative to a redox buffer assemblage (figure 5.12). Oxygen thermobarometry analysis of upper mantle spinel peridotites, and garnet peridotites from the cratonic lithosphere indicate that oxygen fugacity decreases with depth, and thus reduced redox conditions in the mantle increase with depth (Frost, 1991; Frost & McCammon, 2008); it is believed that this decrease in oxygen fugacity is a result of the increasing effects of pressure controlling Fe3+/Fe2+ equilibria. Based on this relationship, Bellis and Canil (2006) showed that kimberlite melt crystallization experiments under various ƒO2 conditions would yield different Fe3+ contents in groundmass perovskites (CaTiO3). They showed that the partitioning of Fe3+ into perovskite increased in direct response to increased ƒO2, and that this partitioning was independent of the temperature of crystallization, or the bulk Fe content of the melt.  Figure 5.12 – Common redox buffering assemblages for various ranges of oxygen fugacity as a function of temperature at a constant 1 atm (Frost, 1991). MH = magnetite-hematite, NiNiO = nickel-nickel oxide, FMQ = fayalite-magnetite-quartz, IW = wustite-magnetite, QIF = quartz-iron-fayalite. Discussion - Oxygen Fugacity  120   This partitioning was however affected by the Nb content in the melt, with increased Nb resulting in higher Fe3+ and Nb coupled substitution for Ti in the octahedral B-site (Bellis & Canil, 2006). Based on their results, they calibrated an empirical oxybarometer that models the covariation of Fe and Nb in perovskites as a proxy for oxygen fugacity of the kimberlite melt in equilibrium with the perovskite (Formula 1); oxygen fugacity is calculated relative to the nickel-nickel oxide (NiNiO) buffer assemblage and experimental data was reproducible within a 1 log unit margin of error. ∆NNO = [0.50(±0.021) x Nb – Fe(±0.031) + 0.030(±0.001)]/0.004(±0.0002)         (1) In a companion paper, Canil & Bellis (2006) studied 11 kimberlites from Somerset Island and Lac de Gras, and compiled kimberlite analyses from available literature, and showed that ƒO2 recorded between kimberlites from different provinces, different pipes within a province, and even different phases within a single pipe, can vary by up to three orders of magnitude. They conclude that it’s therefore feasible to differentiate between phases of kimberlite on the basis of identifying microphenocrystal perovskite populations crystallized under different ƒO2 conditions. For groundmass perovskites in the Renard 65 kimberlites, calculated cation proportions of Nb and Fe (Table 3.5) were used to estimate the ƒO2 (∆NNO) of the kimberlite melt, in equilibrium with perovskite at the time of perovskite crystallization. Due to the small crystal sizes, all of the analyses were collected as core analyses. The compositions thus provide the best estimate of the ƒO2 of the unmodified kimberlite melt. ƒO2 values are plotted by geological unit (figure 5.13). The total range of ƒO2 for Renard 65 are then compared with ƒO2 values reported for the cratonic mantle (Woodland & Koch, 2003; McCammon & Kopylova, 2004), other mantle derived magmas (Carmichael & Ghiorso, 1986; Christie et al., 1986; Carmichael, 1991) and kimberlite perovskites obtained from Lac de Gras, Somerset Island and within the literature (modified after Canil & Bellis, 2006) (figure 5.12).  Discussion - Oxygen Fugacity  121  Perovskite in KPKs from Kimb65a are hosted in magmaclasts, where they are often pseudomorphed by Mn-ilmenite (discussed in detail in Sect. 5.9 Perovskite Alteration by CO2-rich Fluids). Where partial crystals are preserved, they record relatively oxidized conditions in the kimberlite melt during crystallization. The two analyses that record more reduced conditions (-4.5 ∆NNO) could reflect a broader population not represented due to limited sampling, two generations or populations of perovskite, or perhaps the effects of subsolidus alteration processes. Perovskite in HK and HKt from Kimb65b are distributed throughout the interstitial groundmass and record a wide range of conditions from -0.5 to -4 ∆NNO with much of the crystallization occurring over -2 to -3 ∆NNO. Perovskite in KPKt rock types from Kimb65d are hosted in magmaclasts, and they record the most reduced conditions ranging from -2 to -4.5 ∆NNO with the majority of crystallization taking occurring over -3.5 to -4.5 ∆NNO.  These data suggest that despite a fair degree of overlap, the bulk of groundmass and magmaclast perovskites in each of the three geological units’ record crystallization under distinctly different ƒO2 conditions. The broad range of ƒO2 values for each geological unit is likely indicative of systematic changes in ƒO2 prior to and during emplacement (and to some extent during late-stage perovskite crystallization).  Factors reducing ƒO2 and driving more oxidized conditions during ascent and emplacement include crystallization, decompression and volatile degassing (Canil & Bellis, 2006). When Comparing ƒO2 estimates in identifying separate phases of kimberlite additional consideration must be paid to the possible effects that emplacement related processes may have on the late-stage growth of perovskite. Discussion - Oxygen Fugacity  122   Figure 5.13 – Oxygen fugacity ƒO2 expressed as log units relative to the NiNiO buffer for the geological units Kimb65a, Kimb65b and Kimb65d using the perovskite oxygen barometer of Bellis & Canil (2006). The margin of error on the oxygen barometer is 1 log unit. The Renard 65 ƒO2 values are compared with reported ƒO2 values for the cratonic mantle, lamprophyres and minettes, W. Mexico basalt and andesite, Kilauea and MORB, Lac de Gras and Somerset Island kimberlites, and a compilation of analyses from published literature (modified from Canil & Bellis, 2006). Discussion - Oxygen Fugacity  123  It may initially be tempting to conclude that the more reduced conditions recorded by perovskite in KPKt rock types in Kimb65d compared to the more oxidized conditions recorded by HK/HKt rock types in Kimb65b (figure 5.13) would preclude the possibility that these rock types represent a single phase of kimberlite; if processes such as crystallization and volatile degassing are more enhanced in the production of KPK rock types, one would expect higher ƒO2 and more oxidized conditions recorded in these rock types. However, it is important to recognize at what point perovskite growth ceases during the production of these rock types, and thus to what extent it records the late changes in the ƒO2 of the magma. KPK rock types in general are characterized by microlitic textures and seemingly weakly reacted crustal xenoliths and are likely to have cooled much faster than HK rock types. This rapid cooling would result in solidification of magmaclasts and termination of perovskite growth at an earlier time than in HK rock types of the same phase of kimberlite which cooled more slowly and were able to record the late-stage changes in the ƒO2 of the melt. The morphological features of perovskite in magmaclasts in KPKt rock types in Kimb65b and in the groundmass of HK and HKt rock types in Kimb65b are consistent with such a scenario; groundmass perovskite in Kimb65b (figure 5.14A) is larger, displays irregular compositional zoning on the outer margins of the crystals, and in these outer zones are common partial inclusions of ulvöspinel. The irregular compositional zoning and partial inclusion of ulvöspinel (which form late rims on earlier crystallized chromite in both geological units) suggest late-stage growth from a contaminated and residual kimberlite melt. These features are not recorded in perovskite in magmaclasts in KPKt rock types in Kimb65b presumably because perovskite growth was terminated by the solidification of magmaclasts. Therefore, the differences in calculated ƒO2 for Kimb65b and Kimb65d do not preclude the possibility that these geological units represent a single phase of kimberlite. Discussion - Oxygen Fugacity  124   Figure 5.14 – (A) Perovskite hosted in the interstitial groundmass of HK and HKt rock types in Kimb65b. Crystals are generally larger than those in Kimb65b, they display irregular compositional zoning along the crystal margins, and in these zones are often partial inclusions of ulvöspinel. (B) Perovskite hosted in the magmaclasts of KPKt rock types in Kimb65d are generally smaller and do not display compositional zoning or partial inclusions of ulvöspinel.   A more meaningful comparison of ƒO2 can be made between KPK/KPKt rock types in Kimb65a and KPKt rock types in Kimb65d as the effects of emplacement related processes would be reasonably similar in these comparable rock types. Perovskites in KPK/KPKt rock types in Kimb65a record significantly more oxidized conditions during crystallization than do perovskites in KPKt rock types in Kimb65d (figure 5.13) and these differences are likely to reflect the different intrinsic ƒO2 values of each of the kimberlite magmas, supporting their identification as separate phases of kimberlite. This conclusion would be overwhelmingly supported by a large number of differences in the features of kimberlite components (Table 3.1) in these rock types, and sharp cross-cutting contacts between these units (figure 3.12). The total range of ƒO2 values calculated for the Renard 65 kimberlites -0.5 to -5 ∆NNO are consistent with the reduced conditions in the lithospheric mantle (Woodland & Koch, 2003; McCammon & Kopylova, 2004) and show broad overlap with ƒO2 conditions reported for the majority of kimberlite occurrences (compiled in Canil & Bellis, 2006). This may initially seem problematic, as kimberlite magmas crystallize carbonate, the stability of which is defined by the C-H-O fluid equilibria above 0 to -2 ∆NNO assuming XCO2 = 0.5 in the fluid (Canil & Bellis, 2006); thus perovskites in Renard 65, Somerset Island, and most of the reported literature record growth from magmas with ƒO2 too low to stabilize carbonates. However, while kimberlites are carbonate rich rocks, carbonate does not crystallize in equilibrium with perovskite. Discussion - Geological Modelling of the Renard 65 Pipe  125  Carbonate in coherent rock types crystallizes interstitially to earlier formed groundmass minerals including spinel, perovskite, phlogopite, and apatite and is generally absent in KPK rock types where it is presumed to be lost during the exsolution of magmatic volatiles. It is also observed as a replacement product with Mn-ilmenite where perovskite has been altered by deuteric CO2-rich fluids. Thus, the textural relationships between carbonate and perovskite in Renard 65 are consistent with the requirement that carbonate cannot crystallize from a kimberlite magma until an ƒO2 of 0 to -2 ∆NNO is reached which is significantly higher than the ƒO2 recorded during perovskite crystallization. Thus we can postulate an oxidation during crystallization of the kimberlite, from -0.5 to -5 ∆NNO (conditions of perovskite precipitation) to 0 to -2 ∆NNO (conditions of carbonate precipitation). 5.11 Geological Modelling of the Renard 65 Pipe 5.11.1 Identification of Kimberlite Phases for Geological Modelling A phase of kimberlite encompasses all of the texturally and mineralogically variable rock types or solidified products of a single kimberlite magma during a single magmatic event (Scott Smith & Smith, 2009). Identifying a phase of kimberlite involves first characterizing the features of the components of the kimberlite which were equilibrated or crystallized prior to emplacement, and tracking the consistency in these components through the various rock types that result from emplacement related processes. These would include characteristics such as the relative modal abundances of early crystallized groundmass minerals (i.e. spinel, perovskite), the size distribution of olivine macrocrysts, the chemistry of ‘indicator’ mantle xenocrysts (i.e. garnet, ilmenite, chromite), and the macrodiamond grade (Scott Smith & Smith, 2009). Each of these features are meaningful because they are based either on the chemistry of the kimberlite melt, or the grain size characteristics of the entrained mantle xenoliths, or the chemistry and thus pressures and temperatures of equilibration of mantle xenocrysts, or the size frequency distribution(s) of the entrained diamond population(s) that grew within the mantle xenocrysts. All of these features in theory should remain more or less consistent during emplacement, and be identifiable within Discussion - Geological Modelling of the Renard 65 Pipe  126  different textural rock types derived from the same magma; any changes in these features must be justifiable through emplacement or post-emplacement related processes (Scott Smith & Smith, 2009). 5.11.2 Phases of Kimberlite in Renard 65 The kimberlites in Renard 65 were classified into four kimberlite geological units on the basis of their kimberlite component features (Table 3.1), consistent with classifications by Stornoway Diamond Corporation geological staff. Geological units were designated on the basis of (i) macroscopic colour and rock texture, (ii) olivine macrocryst size and abundance characteristics, and (iii) groundmass modal mineralogy. The four kimberlite geological units include KPK and KPKt rock types in Kimb65a (green-blue on figure 1.8), HK and HKt rock types in Kimb65b (brown), HK rock types in Kimb65c, and HKt and KPKt rock types in Kimb65d (orange-brown). HK rock types in Kimb65c were not extensively sampled for the purposes of this research and so they will simply be treated as a single separate phase of kimberlite. In the following sub-sections, we will discuss the characteristic features of each geological unit, speculate on the effects of emplacement related processes in modifying kimberlite component features, and interpret the number of kimberlite phases represented by these geological units. This will allow for final conclusions on the impacts of this research on the geological model of the Renard 65 pipe. 5.11.2.1 Phase A Perhaps the most appropriate place to start is with KPK and KPKt rock types classified into the geological unit Kimb65a, as these rock types display the most striking differences in kimberlite component features compared with rock types in Kimb65b and Kimb65d which are far more similar to each other. Kimb65a defines a distribution of KPK and KPKt rock types in a cylindrical or conical downward-tapering body with steep near-vertical contacts with the surrounding country rock and kimberlite geological units (figure 1.8 & 5.15). There are a number of petrological features which distinguish these kimberlites from HK, HKt and KPKt rock types in Kimb65b and Kimb65d and allow for their classification as a distinct phase of Kimberlite: Discussion - Geological Modelling of the Renard 65 Pipe  127  1) Crustal xenoliths are well preserved displaying their original mineralogy and textures with eckermannite, chlorite and pyrite forming the reaction mineral assemblage. Monomineralic coronas containing diopside or phlogopite are never observed on crustal xenoliths in these rock types.  2) Pseudomorphed olivine macrocrysts range in size from 1 – 4 mm and less commonly up to 8 mm in size whereas olivine macrocrysts in HK and HKt rock types Kimb65b range from 2 – 8 mm and less commonly up to 16 mm in size, and olivine macrocrysts in KPKt rock types in Kimb65d range from 4 – 8 mm in size. 3) Olivine macrocrysts and microcrysts are entirely pseudomorphed by calcite which is variably overprinted by serpentine in contrast to olivine in HK, HKt and KPKt rock types in Kimb65d which is either fresh, partially or fully pseudomorphed by serpentine.  4) Groundmass perovskite hosted in magmaclasts is nearly entirely altered and replaced by assemblages of Mn-ilmenite ± titanite ± pyrite whereas groundmass perovskite in HK, HKt, and KPKt rock types of Kimb65d are pristinely preserved.  5) Groundmass phlogopite is significantly less abundant than in HK, HKt and KPKt rock types of Kimb65b and Kimb65d. These features support the interpretation that KPK and KPKt rock types of Kimb65a represent the solidified products of a relatively anhydrous kimberlite melt which will now be referred to as Phase A, crystallizing eckermannite in crustal xenoliths and limited groundmass phlogopite now hosted in magmaclasts. The magma was able to retain significant quantities of dissolved CO2 until relatively shallow depths during emplacement where it was able to alter groundmass perovskite and olivine macrocrysts and microcrysts in the CO2-rich environment. The smaller range of crystal sizes for pseudomorphed olivine macrocrysts compared to rock types in other geological units (Table 3.1) suggests that the mantle peridotites entrained and disaggregated by this kimberlite melt were characterized by a smaller average  Discussion - Geological Modelling of the Renard 65 Pipe  128   Figure 5.15 – Sharp planar cross-cutting contacts are observed between KPK and KPKt rock types in Kimb65a and HK/HKt rock types in Kimb65b and Kimb65d in drill-core (DC). olivine size distribution than mantle peridotites sampled by kimberlites in Kimb65b or Kimb65d. In addition to these features, KPK and KPKt rock types in Kimb65a are often observed to have sharp planar contacts crosscut by HK and HKt rock types in Kimb65b (figure 5.15). 5.11.2.2 Phase B The features of the HK, HKt, and KPKt rock types in Kimb65b and Kimb65d are quite similar, making it a very challenging process to determine whether or not these two geological units represent a single phase of kimberlite or two separate phases of kimberlite. HK and HKt rock types in Kimb65b are distributed throughout a near-vertical cylindrical body generally surrounded or enveloped by the larger cylindrical body of KPKt rock types in Kimb65d (figure 1.8). Their distribution in the pipe suggests that either (i) they represent separate phases of kimberlite and Kimb65b was emplaced generally within the previously emplaced Kimb65d kimberlites, or (ii) they represent a single phase of kimberlite, with gradational changes in texture from HK in the center of the column to KPKt on what would likely be the chilled margins of the nested pipe. Sharp cross-cutting contacts between rock types in each of these geological units are not observed, and thus the megascopic features and distribution of these rock types does not aid in constraining the number of phases that may be represented. Furthermore, macrodiamond grades for each of these geological units (as determined by Stornoway Diamond Corporation) are very Discussion - Geological Modelling of the Renard 65 Pipe  129  similar, meaning that one or more phases of kimberlite cannot be differentiated on the basis of macrodiamond populations. We therefore assess only the petrological features of the kimberlites in these units in identifying the number of phases which they represent. The features of kimberlite components in Kimb65b and Kimb65d are broadly consistent (Table 3.1); they display nearly identical groundmass modal mineral assemblages, interclast matrix assemblages, olivine pseudomorph products, and crustal xenolith reaction styles and mineralogy. In addition, many of the contrasting features including higher crustal xenolith abundances, lower olivine abundances, and common spherical magmaclasts in Kimb65d can be attributed to the effects of emplacement related processes and do not necessarily indicate intrinsic differences between melts. There are however, a number of differences that may or may not be attributable to emplacement related processes as we will now discuss: 1) Crustal xenoliths in HK, HKt and KPKt rock types in Kimb65b and Kimb65d show consistency in their mineralogy and texture. In all rock types the cores of crustal xenoliths are replaced by assemblages of pectolite, calcite, and serpentine, rimmed by sequential monomineralic coronas of diopside then phlogopite. The crustal xenoliths differ only in the development of aegirine-diopside zoning where coarse crystals of aegirine penetrating towards the core of the xenoliths form a continuous gradation with the fine-grained diopside corona. Aegirine-diopside zoning becomes increasingly well developed from HK to HKt rock types in Kimb65b and even more-so in KPKt rock types in Kimb65d. 2) Olivine macrocrysts in the 8 – 16 mm size range are common in HK and HKt rock types in Kimb65b but are uncommon in KPKt rock types in Kimb65d (Table 3.1). 3) The compositions of groundmass spinels in KPKt rock types in Kimb65d have significantly higher Mg #’s than do groundmass spinels in HK rock types in Kimb65b (figure 3.21). The Mg #’s of spinels Discussion - Geological Modelling of the Renard 65 Pipe  130  in HKt rock types in Kimb65b display a complete range of compositions intermediate between the above HK and KPKt rock types. 4) The morphological features including crystal sizes, zoning, or inclusions, of primary spinel, perovskite, and phlogopite are slightly different in HK/HKt rock types in Kimb65b compared with KPKt rock types in Kimb65d. Spinel in both units displays either a chromite or magnesiochromite core composition mantled by an ulvöspinel rim. The ulvöspinel rims in Kimb65b are much thicker and do not conform to the outer growth faces of the chromite core (figure 5.16). Perovskites in Kimb65b are larger, and display irregular compositional zoning on the outer crystal margins with common partial inclusions of ulvöspinel, while perovskite in Kimb65b is smaller, does not display compositional zonation and does not contain partial inclusions of ulvöspinel. Phlogopite crystals in Kimb65b are larger and do not display compositional zonation to Al-depleted and Fe-enriched tetraferriphlogopite compositions as do those in Kimb65d.  Figure 5.16 – Morphological differences between groundmass spinel, perovskite and phlogopite in KPKt rock types of Kimb65d compared with HK and HKt rock types in Kimb65b. Groundmass spinel in Kimb65b displays much more extensively developed ulvöspinel mantles that do not conform to the chromite core growth faces as observed for spinels in Kimb65d. Groundmass perovskite in Kimb65b is larger, displays irregular compositional zoning and contains partial ulvöspinel inclusions while perovskite in Kimb65d is smaller, does not display zoning, and does not contain partial inclusions of ulvöspinel. Groundmass phlogopite in Kimb65b is larger and does not display zoning while those in Kimb65d are smaller and display zoning from phlogopite to tetraferriphlogopite rims. Discussion - Geological Modelling of the Renard 65 Pipe  131  We now consider whether or not these differences can reasonably be attributed to emplacement-related processes affecting a single phase of kimberlite, or whether they reflect the presence of two separate batches of kimberlite melt. The more conservative approach would be to suggest that the differences in these components cannot be reliably proven to result from emplacement and subsolidus processes and that they must represent separate phases of kimberlite. In this scenario, the different olivine macrocryst size distributions in the kimberlites would reflect a difference in the olivine size distribution of entrained mantle xenoliths. The high and low Mg # end-member compositions of groundmass spinels would represent crystallization from different melts, and presumably the HKt rock types with a complete intermediate range of compositions would reflect the effects of mixing between the magma of Kimb65b and the solidified rocks of Kimb65d through which it was emplaced. The differences in morphological features of groundmass spinel, perovskite and phlogopite would simply be attributed to crystallization from different batches of melt, and the development of aegirine-diopside zoning in the coronas or crustal xenoliths could simply be a reflection of emplacement and post-emplacement processes resulting in higher Fe3+/FeT ratio of Kimb65b (figure 5.13). The reason that this interpretation is problematic is that it emphasizes what amount to very minor differences between two otherwise incredibly similar sets of rock types (Table 3.1). We thus instead attempt to address these differences as a result of emplacement-related processes to demonstrate the viability in interpreting these two geological units as a single phase of kimberlite. In Sect. 5.6.2 Reactions Between Crustal Xenoliths in HK, HKt, and KPKt Rock Types in Kimb65b and Kimb65d, it was shown that the development of aegirine-diopside zoning was in response to an increase in the availability of Fe3+, which could be explained to result from Fe oxidation by serpentinizing fluids. HK rock types which display minimal or absent aegirine-diopside zoning show minimal or partial olivine serpentinization; HKt and KPKt rock types which display increasingly well-developed aegirine-diopside zoning show the most extensive olivine serpentinization. The increased a(Fe3+) required to produce Discussion - Geological Modelling of the Renard 65 Pipe  132  aegirine zoning can be explained by subsolidus exchange of Mg for Fe in groundmass spinels, producing a trend of increasing Mg #’s from spinels in HK rock types to KPKt rock types, with spinels in HKt rock types preserving the intermediate compositions of this subsolidus exchange process. Thus, the development of aegirine-diopside zoning, and the compositional differences in groundmass spinels between Kimb65b and Kimb65d can arguably have resulted from the variable effects of emplacement and post-emplacement processes on a single batch of kimberlite melt. The differences in morphological features of groundmass spinels, perovskites, and phlogopites as described above (figure 5.16) may also have resulted from the effects of emplacement-related processes. As described in Sect. 5.10 Oxygen Fugacity, KPK rock types are likely to have formed from a rapidly cooling magma, as indicated by their abundance of microlitic textures, while HK rock types are likely to have cooled much more slowly as evidenced by their more crystalline groundmass. As a result, the late-stage growth of groundmass minerals hosted within magmaclasts will likely have been terminated earlier than for groundmass minerals hosted within more coherent rock types. It can therefore be argued that the larger crystal sizes for perovskite and phlogopite in Kimb65b compared to Kimb65d (figure 5.16) are a result of slower cooling; this would not and does not affect groundmass spinels which ceased crystallization prior to emplacement. The development of thicker ulvöspinel mantles on chromite or magnesiochromite cores in Kimb65b (figure 5.16), and the irregular compositional zoning and partial inclusions of spinel for perovskites in Kimb65b (figure 5.16), are both indicative of prolonged re-equilibration with or crystallization from a contaminated kimberlite magma. Thus, emplacement-related processes can reasonably account for the majority of observed mineralogical, textural or compositional differences in kimberlite components in Kimb65b and Kimb65d. The only feature that cannot be accounted for is the slightly smaller olivine size distribution in Kimb65d. However, all other aspects of the kimberlite components in the geological units Kimb65b and Kimb65d are either shown to be nearly identical (style of crustal xenolith reaction, groundmass modal mineralogy, interclast matrix composition), and where dissimilar their differences can be attributed to the effects of emplacement-Discussion - Geological Modelling of the Renard 65 Pipe  133  related and/or subsolidus processes (extent of aegirine-diopside zoning, groundmass spinel chemistry, textural features of groundmass minerals). It is therefore reasonable to conclude that HK and HKt rock types in Kimb65b and KPKt rock types in Kimb65d represent a single batch of kimberlite melt and single phase, which will now be referred to as Phase B. This conclusion is, by the way, consistent with the nearly identical macrodiamond grades for these two geological units independently determined by Stornoway Diamond Corporation (Bagnell et al., 2013). 5.11.3 The Emplacement of Renard 65 Kimberlites and the Origin of the Interclast Matrix The emplacement process envisioned for the HK, HKt, KPKt and KPK rock types Phases A and B in the Renard 65 pipe is similar to the subvolcanic fluidization model (Clement, 1982; Mitchell, 1986; Field & Scott Smith, 1999; Skinner & Marsh, 2004) for reasons detailed in Sect. 5.2 Kimberlite Emplacement: Processes of Diatreme Formation and Infill. However, this emplacement model is insufficient as it acknowledges only the effects of decompression-induced volatile exsolution, and the short-lived effects of an upwelling degassing front. This is problematic in describing the processes active during the emplacement of the Renard 65 kimberlites for two reasons: 1) Phase A displays KPK/KPKt textures from the present day surface, estimated to represent approximately 1 km of erosion from the original volcanic surface (Birkett et al., 2003), down to 500 m depth in Renard 65; similarly, in Renard 2 KPK/KPKt textures in Kimb2a are present down to over 900 m depth in the pipe. The exsolution of magmatic volatiles in response to decompression should occur at a certain depth and cannot explain the development of these features over a 500 m to 1 km interval. 2) Phase B displays a complete range of HK, HKt and KPKt rock types over a relatively short interval (< 100 m) but these gradations are preserved laterally and not vertically, which cannot explain by a buoyant and upwelling volatile degassing front. Interestingly, similar non-Discussion - Geological Modelling of the Renard 65 Pipe  134  horizontal transition zones are preserved in the Hearne pipe at Gahcho Kue (Hetman et al., 2004). For these reasons, we argue that the development of transitional and KPK textures in Renard 65 cannot be attributed solely to decompression and the exsolution of magmatic volatiles. If volatile exsolution alone was capable of producing KPK rock types, then similar assemblages of diopside + phlogopite + serpentine should be observed within FPK rock types that have also undergone degassing and fragmentation (Scott Smith, 2008), but these assemblages are not observed in FPK rock types. In Sect. 5.6.2 Reactions Between Crustal Xenoliths in HK, HKt, and KPKt Rock Types in Kimb65b and Kimb65d, it was shown that the hybrid groundmass assemblage diopside + phlogopite + serpentine could result from contamination of the kimberlite melt by Si from crustal xenolith assimilation, shifting the stable groundmass assemblage of phlogopite + calcite + serpentine to diopside + phlogopite + serpentine and exsolving CO2. We therefore propose that the development of this hybrid groundmass assemblage (commonly referred to as the interclast matrix assemblage) is a function of the distribution of crustal xenoliths. The geological unit summary for kimberlites in Renard 65 show a clear correlation between increasing crustal xenolith modal abundances and increasingly transitional and KPK textures (Table 3.1). Similar correlations between crustal xenolith modal abundance and KPK textures are also observed in other KPK pipes from Renard and Gahcho Kue including Renard 3, Tuzo, Hearne, and 5034 (figure 5.17). Discussion - Geological Modelling of the Renard 65 Pipe  135   Figure 5.17 – Correlations between crustal xenolith modal abundances and kimberlite rock textures for kimberlites in multiple pipes from Gahcho Kue (Tuzo, Hearne, 5034) and Renard (Renard 2 & Renard 65). Rock types with transitional to KPK textures in all pipes are associated with increased crustal xenolith modal abundances. Data for the Gahcho Kue pipes Tuzo, Hearne and 5034 are from drill hole summaries in Hetman et al. (2004), and data for the Renard pipes are from drill hole summaries in this research (Renard 65) and geological unit summaries (Renard 3) in Muntener & Scott Smith (2013). This model of KPK emplacement explains the presence of KPK/KPKt rock textures over a 500 m interval in Phase A of Renard 65, and also for the gradational lateral change in texture from HK to HKt to KPKt in Phase B of Renard 65. Such features by the way are not unique to Renard 65; similar and even larger 900 m (depth) continuous intervals of KPK/KPKt rock types are observed in Renard 2, and non-horizontal gradations from HK through transitional to KPK rock types within a single phase of kimberlite were also described for the Hearne pipe at Gahcho Kue (figure 12 from Hetman et al., 2004). This model Discussion - Contrasting Texture and Mineralogy of KPK and FPK Pipes  136  thus accounts for the features of not only the Renard 65 kimberlites, but other well documented KPK pipes. 5.11.4 Renard 65 Geological Model By characterizing the effects of subsolidus emplacement related processes on the mineralogy, textures and compositions of kimberlite components in Renard 65, the geological units Kimb65a, Kimb65b and Kimb65d are reclassified into kimberlite phases A and B respectively (figure 5.18). Providing improved constraints for the identification of kimberlite phases allow for an improved degree of confidence in modelling the spatial distribution of diamond grades throughout the Renard 65 pipe.  Figure 5.18 – Reclassification of the geological units Kimb65a, Kimb65b and Kimb65d in the Renard 65 geological model (figure 5.3) to an updated geological model showing the distribution of kimberlite phases A and B. 5.12 Contrasting Texture and Mineralogy of KPK and FPK Pipes It was proposed by Scott Smith (2008a) that perhaps one of the most influential factors in producing different classes of kimberlite pipes was the host country rock lithology; FPK pipes which typically display high width-to-depth ratios in pipe morphology are correlated with poorly consolidated sedimentary host lithologies, while KPK pipes which typically display steep-sided pipes are correlated with competent basement rock lithologies (figure 5.19). Based on this observed correlation across more than 800 localities,  Discussion - Contrasting Texture and Mineralogy of KPK and FPK Pipes  137   Figure 5.19 - Schematic representation of the geology of three main kimberlite types after (modified after Scott Smith, 2008a). Fort-à-la-Corne type pyroclastic kimberlites (FPK) are associated with poorly consolidated sedimentary lithologies, and form high width to depth ratio pipes infilled with pyroclastic kimberlite. Lac-de-Gras type resedimented volcaniclastic kimberlites (RVK) are correlated with minor poorly consolidated sediments overlying more competent basement rock, forming steep-sided pipes with flared wide craters, infilled with bedded and graded sequences of resedimented volcaniclastic kimberlite with entrained surface material. Kimberley-type pyroclastic kimberlites (KPK), consistent with the majority of South African kimberlites, are emplaced into competent basement rocks, forming steep-sided pipes containing volcaniclastic kimberlite with transitional textures to coherent kimberlite. Scott Smith (2008a) suggests the structural controls imparted by the country rock lithology could affect the exsolution of magmatic volatiles during explosive kimberlite eruptions, resulting in pipes with different morphologies, textures and mineralogy. The current research on the Renard 65 pipe however, demonstrates that magmatic volatile exsolution alone cannot explain the distribution of rock types in KPK pipes; while we agree with Scott Smith (2008a) that the host rock lithology will affect the morphology of the kimberlite pipe (and probably the extent to which crustal xenoliths become entrained in the kimberlite melt), we argue that the development of transitional to KPK textural rock types is a function of the distribution of silicic crustal xenoliths and the extent to which they become assimilated by the kimberlite Discussion - Contrasting Texture and Mineralogy of KPK and FPK Pipes  138  melt. Inversely, the “overall lack of xenoliths” in FPK rock types (Scott Smith, 2008a) may provide a better explanation for the differences in the observed textures and mineralogies of the associated rock types in FPK pipes compared with KPK pipes. Conclusions  139  6. Conclusions 1. The Renard 65 pipe is classified as a Kimberley-type pyroclastic kimberlite pipe containing hypabyssal kimberlite (HK), transitional hypabyssal kimberlite (HKt), transitional Kimberley-type pyroclastic kimberlite (KPKt) and Kimberley-type pyroclastic kimberlite (KPK) rock types. The geological unit ‘Kimb65a’ includes KPK/KPKt rock types and is mineralogically classified as an apatite-phlogopite kimberlite. The geological unit ‘Kimb65b’ includes HK/HKt rock types and is mineralogically classified as a serpentine-phlogopite kimberlite. The geological unit ‘Kimb65c’ includes HK rock types from cross-cutting hypabyssal dykes and is mineralogically classified as a phlogopite-calcite-serpentine kimberlite. The geological unit ‘Kimb65d’ includes HKt/KPKt rock types and is mineralogically classified as a phlogopite kimberlite. 2. Granitoid and gneissic crustal xenoliths are present in all geological units displaying different modal abundances and styles of reaction. Crustal xenoliths in Kimb65a range in modal abundance from 40 to 90 % and display minor reaction mineral assemblages including amphibole + chlorite + pyrite. Crustal xenoliths in Kimb65b and Kimb65d range in modal abundance from 15 – 30 % and 20 – 40 % respectively, and both display reaction mineral assemblages composed of pectolite + calcite + serpentine with monomineralic coronas of diopside and phlogopite; crustal xenoliths in Kimb65d display more developed aegirine-diopside zoning in diopside coronas. 3. Kimberlites in Kimb65a are interpreted to represent a single phase of kimberlite (Phase A) distinguished from kimberlites in Kimb65b and Kimb65d based on distinct differences in their kimberlite component features that include style of crustal xenolith reaction, groundmass modal mineralogy, olivine macrocryst size distribution, preservation of groundmass perovskite, among others. Kimberlites in Kimb65b and Kimb65d are interpreted to represent a single phase of kimberlite as nearly all of the differences in their kimberlite component characteristics, such as the groundmass spinel chemistry, aegirine-diopside zoning in crustal xenoliths, extent of serpentinization can be attributed to the effects of emplacement and post-emplacement processes changing the redox state. Conclusions  140  4. Crustal xenoliths in Renard 65 were entrained initially at high temperatures where in-situ magmatic reactions characterized the interaction between xenolith and melt, progressing into subsolidus temperatures where autometasomatic reactions dominated. Crustal xenoliths in Phase A record initial reaction temperatures estimated to be between 800 – 1000 °C based on the presence of eckermannite, and subsolidus reaction temperatures of 180 – 350 °C based on the chloritization of biotite with intergrowths of pyrite. Crustal xenoliths in Phase B are partially recrystallized to an assemblage of calcite + pectolite + serpentine as likely a result of in-situ reactions with the kimberlite magma by addition of Mg2+, Ca2+, H2O and CO2 to the xenoliths and migration of Na+, K+, Fe2+, Al3+, and Si4+ out of the xenoliths. Contamination of the kimberlite magma by Si results in reduced solubility of CO2 and consequently the exsolution of magmatic volatiles. The presence of monomineralic rims of diopside ± aegirine and phlogopite on crustal xenoliths in Phase B illustrate the autometasomatic process of diffusion of Si out of the crustal xenoliths and Mg into the crustal xenoliths. 5. The model that attributes the kimberlite textures and matrix mineralogy to hydrothermal alteration cannot account for the pervasive pseudomorphing of olivine by calcite, the alteration of groundmass perovskite to Mn-ilmenite + pyrite + calcite, or the continuous lateral gradation between from HK rock types to KPKt rock types in Phase B. The hydrothermal emplacement model is thus considered inadequate in accounting for the interclast matrix mineralogy and textures of Kimberley-type pyroclastic kimberlites. 6. The model that attributes the kimberlite textures and matrix mineralogy to magmatic exsolution cannot explain the continuous KPK/KPKt textures of Phase A to an estimated depth of 1.5 km below the original surface in the pipe, or the preserved lateral gradations in texture from HK in the center of Phase B to KPKt towards the margins of the body. These features in Renard 65 and similar features in other KPK pipes including Renard 2, Renard 3, Tuzo, Hearne, and 5034 are better explained by the development of transitional to pyroclastic textures in response to contamination by silicic crustal xenoliths. Conclusions  141  7. The contrasting mineralogy and textures of Fort-à-la-Corne-type pyroclastic kimberlites and Kimberley-type pyroclastic kimberlites are suggested to be a consequence of not only the structural controls imparted by the country rock lithology, but the extent of contamination by silicic crustal xenoliths.                References  142  References Ballhaus, C., 1993. Redox states of the lithospheric and asthenospheric upper mantle. Contributions to Mineralogy and Petrology. 114, pp.331-348. Becker, M., Le Roex, A.P., 2006. Geochemistry of South African on- and off-craton, Group-I and Group-II kimberlites: petrogenesis and source region evolution. Journal of Petrology. 47(4), pp. 673-703. Bellis, A.J., Canil, D., 2006. Ferric iron in CaTiO3 perovskite as an oxygen barometer for kimberlite magmas I: experimental calibration. Journal of Petrology. 48(2), pp.219-230. Birkett, T.C., McCandless, T.E., Hood, C.T., 2004. Petrology of the Renard igneous bodies: host rocks for diamond in the northern Otish Mountains region, Quebec. Lithos. 76, pp. 475-490. Blank, J.G., Brooker, R.A., 1994. Experimental studies of carbon dioxide in silicate melts; solubility, speciation, and stable carbon isotope behavior. Reviews in Mineralogy and Geochemistry. 30, pp. 157-186. Bagnell, P., Bedell, P., Bertrand, V.J., Brummer, R., Farrow, D., Gagnon, C., Gignac, L.P., Gormely, L., Magnan, M., St-Onge, J.F., 2013. Renard Diamond Project NI 43-101 Technical Report. Unpublished report for Stornoway Diamond Corporation, Longueuil, Quebec. 296p. Boctor, N.Z., Boyd, F.R., 1980. Oxide minerals in the Liqhobong kimberlite, Lesotho. American Mineralogist. 65, pp. 631-628. Boctor, N.Z., Boyd, F.R., 1982. Petrology of kimberlites from the DeBruyn and Martin Mine, Bellsbank, South Africa. American Mineralogist. 67, pp. 917-925. Bowen, N.L., 1928. The evolution of the igneous rocks. Princeton University Press, Princeton. Pp. 334. Brett, R.C., Russell, J.K., Andrews, G.D.M., Jones, T.J., 2015. The ascent of kimberlite: insights from olivine. Earth and Planetary Science Letters. 424, pp. 119-131. Brey, G., Ryabchikov, I., 1994. Carbon dioxide in strongly undersaturated melts and origin of kimberlitic magmas. Abhandlungen Journal of Mineralogy and Geochemistry. H10, pp. 449-463. Brooker, R., Kohn, S., Holloway, J., McMillan, P., 2001. Structural controls on the solubility of CO2 in silicate melts: part I: bulk solubility data. Chemical Geology. 174, pp. 225-239. Brooker, R., Sparks, R.S.J., Kavanagh, J.L., Field, M., 2011. The volatile content of hypabyssal kimberlite magmas: some constraints from experiments on natural rock compositions. Bulletin of Volcanology. 73, 959-981. Brown, R.J., Buse, B., Sparks, R.S.J., Field, M., 2008. On the welding of pyroclasts from very low-viscosity magmas: examples from kimberlite volcanoes. The Journal of Geology. 116, pp. 354-374. Brown, R.J., Manya, S., Buisman, I., Fontana, G., Field, M., Mac Niocaill, C., Sparks, R.S.J., Stuart, F.M., 2012. Eruption of kimberlite magmas: physical volcanology, geomorphology, and age of the References  143  youngest kimberlite volcanoes known on earth (the Upper Pleistocene/Holocene Igwisi Hills volcanoes, Tanzania). Bulletin of Volcanology. 74, pp. 1621-1643. Buse, B., Schumacher, J.C., Sparks, R.S.J., Field, M., 2010. Growth of bultfonteinite and hydrogarnet in metasomatised basalt xenoliths in the B/K9 kimberlite, Damtshaa, Botswana: insights into hydrothermal metamorphism in kimberlite pipes. Contributions to Mineralogy and Petrology. 160, pp. 533-550. Carmichael, I.S.E., 1991. The redox states of basic and silicic magmas: a reflection of their source regions? Contributions to mineralogy and Petrology. 106, pp. 129-141. Carmichael I.S.E., Ghiorso, M.S., 1986. Oxidation-reduction relations in basic magma: a case for homogeneous equilibria. Earth and Planetary Science Letters. 78, pp. 200-210. Caro, G., Kopylova, M.G., Creaser, R.A., 2004. The hypabyssal 5034 kimberlite of the Gahcho Kue cluster, southeastern Slave craton, Northwest Territories, Canada: A granite-contaminated Group-I kimberlite. The Canadian Mineralogist. 42, pp. 183-207. Canil, D., Bellis, A.J., 2006. Ferric iron in CaTiO3 perovskite as an oxygen barometer for kimberlite magmas II: Applications. Journal of Petrology. 48, pp. 231-252. Cas, R.A.F., Porritt, L., Pittari, A., Hayman, P., 2008a. A new approach to kimberlite facies terminology using a revised general approach to the nomenclature of all volcanic rocks and deposits: descriptive to genetic. Journal of Volcanology and Geothermal Research. 174, pp. 226-240. Cas, R.A.F., Hayman, P., Pittari, A., Porrit, L., 2008b. Some major problems with existing models and terminology associated with kimberlite pipes from a volcanological perspective, and some suggestions. Journal of Volcanology and Geothermal Research. 174, pp. 209-225. Chakhmouradian A.R., Mitchell R.H., 2000. Occurrence, alteration patterns and compositional variation of perovskite in kimberlites. Canadian Mineralogist. 38, pp. 978-994. Chakhmouradian A.R., Mitchell R.H., 2001. Three compositional varieties of perovskite from kimberlites of the Lac de Gras field (Northwest Territories, Canada). Mineralogical Magazine 65(1), pp.133-148. Chakhmouradian A.R., 2002. Strontium-apatite: new occurrences, and the extent of Sr-for-Ca substitution in apatite-group minerals. The Canadian Mineralogist. 40, pp. 121-136. Christie, D.M., Carmichael, I.S.E., Langmuir, C.H., 1986. Oxidation states of mid-ocean ridge basalt glasses. Earth and Planetary Science Letters. 79, pp. 397-411. Clement, C.R., 1975. The emplacement of some diatreme facies kimberlites. Physics and chemistry of the earth. 9, pp. 123-136. References  144  Clement, C.R., Skinner, E.M.W., 1979. A textural-genetic classification of kimberlite rocks. Kimberlite Symposium II, Cambridge, Extended Abstract. Clement, C.R., 1982. A comparative geological study of some major kimberlite pipes in the Northern Cape and Orange Free State. Ph.D. thesis (2 vols), University of Cape Town, South Africa. Clement, C.R., Skinner, E.M.W., Scott Smith, B.H., 1984. Kimberlite redefined. Journal of Geology. 92, pp.223-228. Clement, C.R., Skinner, E.M.W., 1985. A textural-genetic classification of kimberlites. Transactions of the Geological Society of South Africa. 88, pp. 403-409. Clement, C.R., Reid, A.M., 1989. The origin of kimberlite pipes: an interpretation based on a synthesis of geological features displayed by southern African occurrences. In: Ross et al. 1, pp. 632-646. Cloos, H., 1941. Bau and Tatigkeit von Tuffschloten. Untersuchungen an den Schwabischen Vulkanen. Geologische Rundschau. 32, pp. 709-800. Coleman, R.G., 1971. Petrologic and geophysical nature of serpentines. Geological Society of America Bulletin. 82, pp. 879-918. Dawson, J.B., 1971. Advances in kimberlite geology. Earth Science Reviews. 7, pp. 187-214. Dawson, J.B., 1980. Kimberlites and their xenoliths. Springer, Berlin. pp. 252. Decarreau, A., Petit, S., Vieillard, P., Dabert, N., 2004. Hydrothermal synthesis of aegirine at 200°C. European Journal of Mineralogy. 16, pp. 85-90. DePaolo, D.J., 1981. A neodymium and strontium isotopic study of the mesozoic calc-alkaline granitic batholiths of the sierra Nevada and Peninsular Ranges, California. 86(B11), pp. 10470-10488. Droop, G.T.R., 1987. A general equation for estimating Fe3+ concentrations in ferromagnesian silicates and oxides from microprobe analysis, using stoichiometric criteria.  Mineralogical Magazine. 51, pp. 431-435. Edgar, A.D., Arima, M., Baldwin, D.K., Bell, D.R., Shee, S.R., Skinner, E.M.W., Walker, E.C., 1988. High pressure – high-temperature melting experiments on a SiO2-poor aphanitic kimberlite from the Wesselton mine, Kimberley, South Africa. American Mineralogist. 73, pp. 524-533. Ernst, W.G., 1968. Amphiboles: crystal chemistry, phase relations and occurrence. Springer-Verlag New York Inc. Evans, B.W., 2004. The serpentine multisystem revisited: chrysotile is metastable. International Geology Review. 46, pp. 479-506. Fedortchouk, Y., Canil, D., 2004. Intensive variables in kimberlite magmas, Lac de Gras, Canada, and implications for diamond survival. Journal of Petrology. 45, pp. 1725-1745. References  145  Field, M., Scott Smith, B.H., 1998. Textural and genetic classification schemes for kimberlite: a new perspective. Extended abstracts, 7th International Kimberlite Conference, Cape Town. pp. 214-216. Field. M., Stiefenhofer, J., Robey, J., Kurszlaukis, S., 2008. Kimberlite-hosted diamond deposits of southern Africa: A review. Ore Geology Reviews. 34(1-2), pp. 33-75. Fontana, G., Mac Niocaill, C., Brown, R.J., Sparks, R.S.J., Field, M., 2011. Emplacement temperatures of pyroclastic and volcaniclastic deposits in kimberlite pipes in southern Africa. Bulletin of Volcanology. 73, pp. 1063-1083. Frantz, G.W., Wyllie, P.J. 1967. Experimental Studies in the system CaO-MgO-SiO2-H2O-CO2. P.J. Wyllie (Ed.), Ultramafic and related rocks, John Wiley and Sons, New York. Pp. 323-326. Frost, B.R., 1991. Introduction to oxygen fugacity and its petrologic significance. In: Lindsley, D.H. (ed.) Oxide Minerals: Petrologic and Magnetic Significance. Mineralogical Society of America, Reviews in Mineralogy. 25, pp. 1-10. Gernon, T.M., Brown, R.J., Tait, M.A., Hincks, T.K., 2012. The origin of pelletal lapilli in explosive kimberlite eruptions. Nature Communications. 3:832. Grant, T.B., Milke, R., Wunder, B., 2014. Experimental reactions between olivine and orthopyroxene with phonolite melt: implications for the origins of hydrous amphibole + phlogopite + diopside bearing metasomatic veins. Contributions to Mineralogy and Petrology. 168: 1073. Guillot, B., Sator, N., 2011. Carbon dioxide in silicate melts: a molecular dynamics study. Geochimica et Cosmochimica Acta. 75, pp. 1829-1857. Haggerty, S.E., 1976. Opaque mineral oxides in terrestrial igneous rocks. In: Rumble, D. (Ed.), Oxide Minerals, Mineralogical Society of America: Reviews in Mineralogy. 3, pp. Hg101-Hg300. Haggerty, S.E., 1987. Metasomatic mineral titanates in upper mantle xenoliths. Mantle Xenoliths (P.H. Nixon, ed.) John Wiley & Sons, New York, N.Y, pp. 671-690). Harris, M. Le Roex, A., Class, C., 2004. Geochemistry of the Uintjiesberg kimberlite, South Africa: petrogenesis of an off-craton, Group-I kimberlite. Lithos. 74(3-4), pp. 149-165. Hasterok, D., Chapman, D.S., 2011. Heat production and geotherms for the continental lithosphere. Earth and Planetary Science Letters. 307, pp. 59-70. Hawthorne, J.B., 1975. Model of a Kimberlite Pipe. Physics and Chemistry of the Earth. 9, pp.1-15. Hayman, P.C., Cas, R.A.F., Johnson, M., 2008. The difficulties in distinguishing coherent from fragmental kimberlite: a case study of the Muskox pipe (Northern Slave Province, Nunavut, Canada). Journal of Volcanology and Geothermal Research. 174, pp. 139-151. References  146  Hayman, P.C., Cas, R.A.F., Johnson, M., 2009. Characteristics and alteration origins of matrix minerals in volcaniclastic kimberlite of the Muskox pipe (Nunavut, Canada). Lithos. 112S, pp. 473-487). Hayman, P.C., Cas, R.A.F., 2011. Reconstruction of a multi-vent kimberlite eruption from deposit and host-rock characteristics: Jericho kimberlite, Nunavut, Canada. Journal of Volcanology and Geothermal Research. 200, pp. 201-222. Hetman, C.M., Scott Smith, B.H., Paul, J.L., Winter, F.W., 2004. Geology of the Gahcho Kue kimberlite pipes, NWT, Canada: root to diatreme magmatic transition zones. Proceedings of the 8th International Kimberlite Conference. Lithos. 76, pp. 51-74. Hetman, C.M., 2008. Tuffisitic kimberlite (TK): A Canadian perspective on a distinctive textural variety of kimberlite. Journal of Volcanology and Geothermal Research. 174, pp. 57-67. Hostetler, P.B., Coleman, R.G., Mumpton, F.A., 1966. Brucite in alpine serpentinites. American Mineralogist. 51, pp. 75-98. Humphris, S.E., 1978. Hydrothermal alteration of oceanic basalts by seawater. Geochimica et Cosmochimica Acta. 42, pp. 107-125. Johannes, W. 1984. Beginning of melting in the granite system Qz-Or-Ab-An-H2O. Contributions to Mineralogy and Petrology. 86, pp. 264-273. Kavanagh, J.L., Sparks, R.S.J., 2009. Temperature changes in ascending kimberlite magma. Earth and Planetary Science Letters. 286, pp. 404-413. Kjarsgaard, B.A., 2007. Kimberlite diamond deposits. In; W.D. Goodfellow (Ed.), Mineral Deposits of Canada: A Synthesis of Major Deposit Types, District Metallogeny, the Evolution of Geological Provinces, and Exploration Methods, Geological Association of Canada, Mineral Deposits Division, Special Publication No. 5 (2007), pp. 245-272. Kjarsgaard, B.A., Pearson, D.G., Tappe, S., Nowell, G.M., Dowall, D.P., 2009. Geochemistry of hypabyssal kimberlite from Lac de Gras, Canada: comparisons to a global database and application to the parent magma problem. Lithos. 112, pp. 236-248. Kopylova, M.G., Gurney, J.J., Daniels, L.R.M., 1997a. Mineral inclusions in diamonds from the River Ranch kimberlite, Zimbabwe. Contributions to Mineralogy and Petrology. 129, pp. 366-384. Kopylova, M.G., Rickard, R.S., Kleyenstueber, A., Taylor, W.R., Gurney, J.J., Daniels, L.R.M., 1997b. First occurrence of strontian K-Cr loparite and Cr-chevkinite in diamonds. Russian Geology and Geophysics, pp. 405-420. Lamadrid, H.M., Lamb, W.M., Santosh, M., Bodnar, R.J., 2014. Raman spectroscopic characterization of H2O in CO2-rich fluid inclusions in granulite facies metamorphic rocks. Gondwana Research. 26, pp. 301-310. References  147  Le Bas, M. J., 2008. Fenites associated with carbonatites. The Canadian Mineralogist. 46, pp. 915-932. Le Roex, A.P., Bell, D.R., Davis, P., 2003. Petrogenesis of Group-I kimberlites from Kimberley, South Africa: evidence from bulk-rock geochemistry. Journal of Petrology. 44(12), pp. 2261-2286. Lewis, H.C., 1887. On diamondiferous peridotite and the genesis of the diamond, Geological Magazine. 4, pp. 22-24. Li, G., Peacor, D.R., Essene, E.J., 1998. The formation of sulfides during alteration of biotite to chlorite-corrensite. Clays and Clay Minerals. 46(6), pp. 649-657. Lorenz, V., 1975. Formation of phreatomagmatic maar-diatreme volcanoes and its relevance to kimberlite diatremes. Physics and Chemistry of the Earth. 9, pp. 17-27. Lorenz, V., Kurszlaukis, S., 2006. Root zone processes in the phreatomagmatic pipe emplacement model and consequences for the evolution of maar-diatreme volcanoes. Journal of Volcanology and Geothermal Research. 159, pp. 4-32. Mannard, G.W., 1962. The Singida kimberlite pipes, Tanganyika. Ph.D. thesis, McGill University, Montreal, Canada. McCallum, M.E., Woolsey, T.S., Schumm, S.A., 1976. A fluidization mechanism for subsidence of bedded tuffs in diatremes and related volcanic vents: Bulletin of Volcanology. 39, pp. 1-16. McCallum, M.E., 1985. Experimental evidence for fluidization processes in breccia pipe formation. Economic Geology. 80, pp.1523-1543. McCammon, C., Kopylova, M.G., 2004. A redox profile of the Slave mantle and oxygen fugacity control in the cratonic mantle. Contributions to Mineralogy and Petrology. 148, pp. 55-68. McGetchin, T.R., 1968. The Moses Rock dike: geology, petrology and mode of emplacement of kimberlite bearing breccia dike, San Juan County, Utah. Ph.D. thesis, California Institute of Technology, Pasadena, CA. McLeod, P., Sparks, R.S.J., The dynamics of xenolith assimilation. Contributions to Mineralogy and Petrology. 132, pp. 21-33. McPhie, J., Doyle, M., Allen, R., 1993. Volcanic textures. Centre of ore deposit and exploration studies University of Tasmania, pp. 196. Mitchell, R.H., 1973. Composition of olivine, silica activity and oxygen fugacity in kimberlite. Lithos. 6, pp. 65-81. Mitchell, R.H., 1979. Mineralogy of the Tunraq Kimberlite, Somerset Island, N.W.T. Canada. In: Boyd and Meyer. 1, pp. 161-171. References  148  Mitchell, R.H., 1985. A review of the mineralogy of lamproites. Transactions of the Geological Society of South Africa. 88, pp.411-437. Mitchell, R.H., 1986. Kimberlites: Mineralogy, Geochemistry, and Petrology, New York: Plenum. Mitchell, R.H., Meyer, H.O.A., 1989. Mineralogy of micaceous kimberlites from the New Elands and Star Mines, Orange Free State, South Africa. In Ross et al. (1989) q.v., 1, pp. 83-96. Mitchell, R.H. 1991. Kimberlites and lamproites: primary sources of diamond. Geoscience Canada. 18, pp. 1-16. Mitchell, R.H., Bergman, S.C., 1991. Petrology of Lamproites. Plenum Press, New York.  Mitchell, R.H., 1994. Suggestions for revisions to the terminology of kimberlites and lamprophyres from a genetic viewpoint. In: Meyer and Leonardos. 1, pp. 15-26. Mitchell, R.H., 1995. Kimberlites, orangeites, and related rocks. Plenum Press, New York. pp. 410. Mitchell, R.H., Chakhmouradian, A.R., 1998. Instability of perovskite in a CO2-rich environment: examples from carbonatite and kimberlite. Canadian Mineralogist. 36, pp. 939-952. Mitchell, R.H., Skinner, E.M.W., Scott Smith, B.H., 2009. Tuffisitic kimberlites from the Wesselton Mine, South Africa: Mineralogical characteristics relevant to their formation. Lithos. 112S. pp 452-464. Mitchell, R.H., 2009. Tuffisitic Kimberlites: mineralogical characteristics relevant to their formation. Lithos. 112, pp. 452-464. Mitchell, R.H., 2013. Mineralogy of magmaclasts and interclast matrices of Kimberley-type pyroclastic kimberlites from the Kao, Letseng-la-Terae, Lethlakane and Premier kimberlite pipes in southern Africa. 10th International Kimberlite Conference Final Extended Abstract No. 10IKC-094. Moussallam, Y., Morizet, Y. Massuyeau, M., Laumonier, M., Gaillard, F., 2015. CO2 solubility in kimberlite melts. Chemical Geology. 418, pp. 198-205. Moussallam, Y., Morizet, Y., Gaillard, F., 2016. H2O-CO2 solubility in low SiO2-melts and the unique mode of kimberlite degassing and emplacement. Earth and Planetary Science Letters. 447, pp. 151-160. Muntener, C., Scott Smith, B.H., 2013. Economic Geology of Renard 3, Quebec, Canada: A diamondiferous, multi-phase pipe, infilled with hypabyssal and Tuffisitic kimberlite. Proceedings of the 10th International Kimberlite Conference. 2, pp. 241-256. Nabelek, P.I., Bedard, J.H., Hryciuk, M., Hayes, B., 2013. Short-duration contact metamorphism of calcareous sedimentary rocks by Neoproterozoic Franklin gabbro sills and dykes on Victoria Island, Canada: Journal of Metamorphic Geology. 31, pp. 205-220. Nesbitt, H.W., Bancroft, M.G., Fyfe, W.S., Karkhanis, S.N., Nishijima, A., Shin, S., 1981. Thermodynamic stability and kinetics of perovskite dissolution. Nature. 289, pp. 358-362. References  149  O’Hanley, D.S., 1996. Serpentinites: records of tectonic and petrological history. New York: Oxford University Press, 1996. Otto, J.W., Wyllie, P.J., 1993. Relationships between silicate-melts and carbonate precipitating melts in CaO-MgO-SiO2-CO2-H2O at 2 kbar. Contributions to Mineralogy and Petrology. 48, pp. 343-365. Page, N.J., 1967. Serpentinization at Burro Mountain, California. Contributions to Mineralogy and Petrology. 14, pp. 321-342. Parfitt, E.A., Wilson, L., 2008. Fundamentals of physical volcanology. Malden – Blackwell Publishing Inc. Pell, J., Russell, J.K., Zhang, S., 2015. Kimberlite emplacement temperatures from conodont geothermometry. Earth and Planetary Science Letters. 411, pp. 131-141. Percival, J.A., 2007. Geology and metallogeny of the superior province, Canada. Mineral Deposits of Canada: A Synthesis of Major Deposit-Types, District Metallogeny, the Evolution of Geological Provinces, and Exploration Methods: Geological Association of Canada, Mineral Deposits Division, Special Publication. 5, pp. 903-928. Porritt, L.A., Cas, R.A.F., Crawford, B.B., 2008. In-vent column collapse as an alternative model for massive volcaniclastic kimberlite emplacement: an example from the Fox kimberlite, Ekati Diamond Mine, NWT, Canada. Journal of Volcanology and Geothermal Research. 174(1-3), pp. 90-102. Phillips, D., Harris, J., 2003. The effect of differential mineral compressibility of diamond inclusion barometry. Abstract: 8th International Kimberlite Conference, Victoria, BC, Canada. Pouchou, J.L., Pichoir, F., 1985. PAP (Z) procedure for improved quantitative microanalysis. Microbeam Analysis. 203, pp. 104-106. Price, S.E., Russell, J.K., Kopylova, M.G., 2000. Primitive magma from the Jericho pipe, N.W.T., Canada: constraints on primary kimberlite melt chemistry. Raudsepp, M., Turncock, A.C., Hawthorne, F.C., 1991. Amphibole synthesis at low pressure: what grows and what doesn’t. European Journal of Mineralogy. 3, pp. 983-1004. Russell, J.K., 1987. Crystallization and vesiculation of the 1984 eruption of Mauna Loa. Journal of Geophysical Research: Solid Earth. 92, pp. 13731-13743. Russell, J.K., Porritt, L.A., Lavalee, Y., Dingwell, D.B., 2012. Kimberlite ascent by assimilation-fuelled buoyancy. Nature. 481, pp. 352-356. Smith, C.B., Gurney, J.J., Skinner, E.M.W., Clement, C.R., Ebrahim, N., 1985. Geochemical character of South African kimberlites: a new approach based upon isotopic constraints. Transactions in the Geological Society of South Africa. 88, pp. 267-280. References  150  Scott Smith, B.H., 1983. Further data on the occurrence of pectolite in kimberlite. Mineralogical Magazine. 47, pp. 75-78. Scott Smith, B.H., 2008a. Canadian Kimberlites: geological characteristics relevant to emplacement. Journal of Volcanology and Geothermal Research. 174, pp. 9-19. Scott Smith, B.H., 2008b. The Fort a la Corne Kimberlites, Saskatchewan, Canada: Geology, Emplacement, and Economics. Journal of the Geological Society of India, 71, pp. 11-55. Scott Smith, B.H., Smith, S.C.S., 2009. The economic implications of kimberlite emplacement. Proceedings of the 9th International Kimberlite Conference. Lithos. 112, pp. 10-22. Scott Smith, B.H., Nowicki, T.E., Russell, J.K., Webb, K.J., Mitchell, R.H., Hetman, C.M., Harder, M., Skinner, E.M.W., Robey, Jv.A., 2013. Kimberlite terminology and classification. Proceedings of the 10th International Kimberlite Conference, Special Issue of the Journal of the Geological Society of India. 2, pp. 1-17. Scott Smith, B.H., Nowicki, T.E., Russell, J.K., Webb, K.J., Mitchell, R.H., Hetman, C.M., Harder, M., Skinner, E.M.W., Robey J.V., (In Press) A glossary of kimberlite and related terms.  Skinner, E.M.W., Scott, B.H., 1979. Petrology, mineralogy, and geochemistry of kimberlite and associated lamprophyre dykes near Swartruggens, W. Transvaal, R.S.A. Kimberlite Symposium II, Cambridge, Extended Abstract. Skinner, E.M.W., Clement, C.R., 1979. Mineralogical classification of southern African kimberlites. Boyd FR, Mayer HOA (eds) Kimberlites, diatremes, and diamonds: their geology, petrology and geochemistry. American Geophysical Union, Washington. pp. 129-139. Skinner, E.M.W., Marsh, S.J., 2004. Distinct kimberlite pipe classes with contrasting eruption processes. Lithos. 76, pp. 183-200. Skinner, E.M.W., 2008. The emplacement of class 1 kimberlites. Journal of Volcanology and Geothermal Research. 174, pp. 40-48. Smirnov, G.I., 1959. On the mineralogy of Siberian kimberlites. Trudy Yakut. Fil., Akad. Nauk SSSR (Proc. Yakut. Branch Acad. Sci. USSR), Ser. Geol. 4, 47-73 (in Russian). Smith, J.V., Brennescholtz, R., Dawson, J.B., 1978. Chemistry of micas from kimberlites and xenoliths. I. Micaceous kimberlites. Geochimica et Cosmochimica Acta. 42, pp. 959-971. Smith, C.B., 1983. Pb, Sr, and Nd isotopic evidence for sources of African Cretaceous kimberlites. Nature. 304, pp. 51-54. Sparks, R.S.J., Baker, L., Brown, R.J., Field, M., Schumacher, J., Stripp, G., Walters, A., 2006. Dynamical constraints on kimberlite volcanism. Journal of Volcanology and Geothermal Research. 155, pp. 18-48. References  151  Sparks, R.S.J., Brooker, R.A., Field, M., Kavanagh, J., Schumacher, J.C., Walter, M.J., White, J., 2009. The nature of erupting kimberlite melts. Lithos. 112S (1), pp. 429-438. Sparks, R.S.J., 2013. Kimberlite volcanism. Annual Review of Earth and Planetary Sciences. 41, pp. 497-528. Stripp, G.R., Field, M., Schumacher, J.C., Sparks, R.S.J., Cressey, G., 2006. Post-emplacement serpentinization and related hydrothermal metamorphism in a kimberlite from Venetia, South Africa. Journal of Metamorphic Geology. 24, pp. 515-534. Thayer, T.P., 1966. Serpentinization considered as a constant volume metasomatic process. American Mineralogist. 51, pp. 685-710. Thayer, T.P., 1967. Serpentinization considered as a constant volume metasomatic process: a reply. American Mineralogist. 52, pp. 549-553. Tucker, M.E., 2001. Sedimentary Petrology: An Introduction to the Origin of Sedimentary Rocks, 3rd edition. Blackwell Science, Malden, MA. van Straaten, B.I., Kopylova, M.G., Russell, J.K., Webb, K.J., Scott Smith, B.H., 2007. Discrimination of diamond resource and non-resource domains in the Victor North pyroclastic kimberlite, Canada. Journal of Volcanology and Geothermal Research. 174, pp. 128-138. Wagner, P.A., 1914. The diamond fields of South Africa. Transvaal Leader, Johannesburg South Africa. Walters, A.L., Phillips, J.C., Brown, R.J., Field, M., Gernon, T., Stripp, G., Sparks, R.S.J., 2006. The role of fluidisation in the formation of volcaniclastic kimberlite: grain size observations and experimental investigation. Journal of Volcanology and Geothermal Research. 115, pp. 119-137. Wilcox, R.E., Petrology of Paricutin Volcano, Mexico. United States Geological Survey Bulletin. 954-C, pp. 281-354. Willcox, A., Buisman, I., Sparks, R.S.J., Brown, R.J., Manya, S., Schumacher, J.C., Tuffen, H., Petrology, geochemistry and low-temperature alteration of lavas and pyroclastic rocks of the kimberlitic Igwisi Hills volcanoes, Tanzania. Chemical Geology. 405, pp. 82-101. Wilson, L., Head III, J.W., 2007. An integrated model of kimberlite ascent and eruption. Nature. 447, pp. 53-57. Winter, J.D., 2010. Principles of Igneous and Metamorphic Petrology, Second Edition. Pearson Education. Woodland, A.B., Koch, M., 2003. Variation in oxygen fugacity with depth in the upper mantle beneath Kaapvaal craton, Southern Africa. Earth and Planetary Science Letters. 214, pp. 295-310.    152  Wooley, A.R., Bergman, S., Edgar, A.D., Le Bas, M.J., Mitchell, R.H., Rock, N.M.S., Scott Smith, B.H., 1995. Classification of lamprophyres, lamproites, kimberlites, and the kalsilite-, melilite-, and leucite-bearing rocks. (Recommendations of the IUGS Subcomission on the Systematics of Igneous Rocks). Canadian Mineralogist. 32(2), pp. 175-186. Woolsey, T.S., 1973. Physical modelling of diatreme emplacement: unpublished M.S. thesis, Colorado State University., 92pg. Woolsey, T.S., McCallum, M.E., Schumm, S.A., 1973. Physical modelling of diatreme emplacement by fluidization. Physics and Chemistry of the Earth. 9, pp. 29-42. Yuguchi, T., Sasao, E., Ishibashi, M., Nishiyama, T., 2015. Hydrothermal chloritization processes from biotite in the Toki granite, Central Japan: Temporal variations of the compositions of hydrothermal fluids associated with chloritization. Mineralogical Society of America. 100, pp. 1134-1152.   Appendix A – Renard 65 Core Logs  153  Appendix A – Renard 65 Core Logs See electronic appendices attached.                         Appendix B – Renard 65 Polished Slab Descriptions  154  Appendix B – Renard 65 Polished Slab Descriptions See electronic appendices attached.                        Appendix C – Renard 65 Mineral Chemistry  155  Appendix C – Renard 65 Mineral Chemistry See electronic appendices attached.  

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.24.1-0340671/manifest

Comment

Related Items