T ERMI NAL STAGE OF A DYING RIFT By Andrew LaVigne A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Geological Sciences - Master of Science 2019 ABSTRACT TERMINAL STAGE OF A DYING RIFT By Andrew LaVigne While most continental rifts progress toward rupture and eventual oceanic spreading, in certain circumstances, the rift may fail. Failed rifts provide a window into the transition from continental r ifting to the formation of a passive mar gin , which in successful rifts is occluded by thick post rift sedimentary packages. Among the best - preserved failed rifts is the 1.1 Ga Midcontinent Rift (MCR) in North America. Within the MCR, the final stage of mag matism is preserved on Michipicoten Isla nd . Here I present a geochemical and isotopic study of the Michipicoten Island Formation to probe conditions in the crust and mantle during the final stage. My results show that the volcanic units on Michipicoten Isl and have undergone magma mixing between rhy olitic and basaltic magma , dominating magmatic processes within the crust. During previous eruptive periods in the MCR , during which this observation has been made, the rhyolitic endmember has been interpreted to have experienced significant contributio n f rom the Achaean crust based on som i values in the MCR , indicating that the source of evolved melts in the MCR changes from melting of existing Archean crust to juveni le material . The isotopic data from Michipicoten Island also shows that the depleted mantle is t he single largest contributor of any geochemical reservoir . In the absence of a strong thermal plume component, melting of the depleted mantle requires decompre ssion. The Michipicoten Island Formation was erupted during the geophysically - defined post - rift phase. My results require plate thinning to have continued during this late stage, with the implication that plate deformation persisted even though the crustal structure may not have recorded this deformation. Copyright by A NDREW LAVIGNE 2019 iv This thesis is dedicated to my grandfather Michael Brink Thank you for inspiring me to be become a scientist v ACKNOWLEDGMENTS I would like to thank the National Sc ience Foundation for funding this research. I also would like to thank the Ontario Park Service for allowing us to conduct our research on Michipicoten Island. I am extremely grateful for all the time and effort my advi sor Dr. Tyrone Rooney put into making there busy schedules to help a young scientist. I would like to thank Dr. Val Finlayson and Dr. Jasper Konter for all the hel p they gave me during the isotopic an alyses stage of this research. Finally, I would like to thank my friends and family whose support over these last few years has been invaluable. vi TABLE OF CONTENTS LIST OF TABLES ................................ ................................ ................................ ..................... viii LIST OF FIGURES ................................ ................................ ................................ .................... ix KEY TO ABBREVIATIONS ................................ ................................ ................................ ........ xi Introduction ................................ ................................ ................................ ................................ . 1 Geologic Bac kground ................................ ................................ ................................ .................. 4 G.1 Geological Setting ................................ ................................ ................................ ..... 4 G.2 Tectonic Setting ................................ ................................ ................................ ........ 4 G.3 Geologic E volution of the Midcontinent R ift ................................ ............................... 5 G.3.1 Initiation stage (1115 - 1110 Ma) ................................ ................................ .. 6 G.3.2 Early Stage (1110 - 1105 Ma) ................................ ................................ ...... 6 G.3.3 Hiatus Stage (1105 - 1101 Ma) ................................ ................................ .... 6 G.3.4 Main Stage Volcanism (1101 - 1094 Ma) ................................ ...................... 7 G.3.5 Late Stage (1094 - 1080 Ma ) ................................ ................................ ........ 7 G.4 Geology of Michipicoten Island ................................ ................................ ................. 8 Methods ................................ ................................ ................................ ................................ .... 1 1 M.1 Major and Trace Element Analyses ................................ ................................ ........ 1 1 M.2 Isotopic Analyses ................................ ................................ ................................ .... 11 M.3 Isotopic Age Correction ................................ ................................ .......................... 1 3 M.3.1 Rb - Sr System ................................ ................................ ........................... 1 4 M.3.2 Sm - Nd System ................................ ................................ ......................... 1 5 M.3.3 Lu - Hf System ................................ ................................ ........................... 1 8 M.3.4 U - Th - Pb System ................................ ................................ ....................... 2 1 M.4 Isotopic Mixing ................................ ................................ ................................ ........ 2 2 Results ................................ ................................ ................................ ................................ ...... 2 5 R.1 Petrography ................................ ................................ ................................ ............ 2 5 R.2 Geochemistry ................................ ................................ ................................ .......... 2 7 R.2.1 Major Elements ................................ ................................ ........................ 2 7 R.2.2 Trace elements ................................ ................................ ......................... 2 9 R.3 Isotope geochemistry ................................ ................................ .............................. 3 1 Discussion ................................ ................................ ................................ ................................ 3 4 D.1 Effects of Alteration ................................ ................................ ................................ . 3 4 D.1.1 Major and Trace Elements ................................ ................................ ........ 3 4 D.1.2 Effec ts of Alteration on Radiogenic Isotope Tracers ................................ . 3 5 D.2 Generation of Evolved Magmatism ................................ ................................ ......... 3 8 D.2.1 Fractional Crystallization ................................ ................................ ........... 3 8 D.2.2 Crustal Anatexis ................................ ................................ ....................... 3 9 D.2.3 Liquid Immiscibility ................................ ................................ .................... 40 D.3 T emporal Evolution of the Michipicoten Island Formation ................................ ....... 41 D.4 Temporal C ontext of the Michipicoten Island Formation in the F ramework of MCR M agmatic S tratigraphy ................................ ................................ ................................ ... 4 3 vii D.4.1 Early Stage ................................ ................................ ............................... 4 4 D.4.2 Hiatus Stage ................................ ................................ ............................. 4 4 D.4.3 Main Stage ................................ ................................ ............................... 4 5 D.5 Hybridization of Late Stage Magmatism in the MCR ................................ ............... 4 7 D.6 Mantle S ources of L ate M agmatism in th e MCR ................................ ..................... 4 8 D.7 Implication for Rift Failure ................................ ................................ ....................... 52 Conclusion ................................ ................................ ................................ ................................ 5 3 APPENDICES ................................ ................................ ................................ ........................... 5 5 APPENDIX A FIGURES ................................ ................................ ................................ 5 6 APPENDIX B TABLES ................................ ................................ ................................ .. 8 7 APPENDIX C STANDARDS ................................ ................................ ....................... 102 REFERENCES ................................ ................................ ................................ ....................... 11 3 viii LIST OF TABLES Table 1 : Major Elements Concentrations ................................ ................................ ................... 8 8 Table 2: Trace Elements Concentrations ................................ ................................ .................. 90 Table 3: Measured Isotopic Values ................................ ................................ ........................... 9 9 Table 4: Calculated Isotopic Values ................................ ................................ ........................ 101 Table 5: Major E lement S tandards ................................ ................................ .......................... 1 0 3 Table 6: Trace Element Standards ................................ ................................ ......................... 10 4 ix LIST OF FIGURES Figure 1: Gravity Anomaly Map ................................ ................................ ................................ . 5 7 Figure 2: MCR Stratigraphy ................................ ................................ ................................ ....... 5 8 Figure 3 : G eologic Map of Michipicoten Island ................................ ................................ .......... 5 9 Figure 4: Photomicrographs ................................ ................................ ................................ ...... 60 Figure 5: Classification Diagrams ................................ ................................ .............................. 61 Figure 6: Major Elements Diagrams ................................ ................................ .......................... 6 3 Figure 7 : Flow Evo lution Diagrams ................................ ................................ ............................ 6 5 Figure 8: Large Ion Lithophile Element Diagrams ................................ ................................ ...... 6 6 Figure 9: First Row Transition Elements Diagrams ................................ ................................ ... 6 7 Figure 10: High Field Strength Element Diagrams ................................ ................................ .... 6 8 Figure 11: Primitive Mantle Normalized Trace Element Diagrams ................................ ............. 70 Figure 12: Chondrite Normalized Rare Earth Element Diagrams ................................ .............. 72 i i Digram ................................ ................................ ................................ ... 7 4 Figure 14: ( 87 Sr/ 86 Sr) i i Diagram ................................ ................................ ....................... 7 5 Figure 15: Pb Isotope Diagrams ................................ ................................ ................................ 7 6 Figure 16: Isocon D iagrams ................................ ................................ ................................ ...... 7 8 Figure 17: Isochron Diagrams ................................ ................................ ................................ .. 7 9 Figure 18: MELTS Models ................................ ................................ ................................ ........ 80 Figure 19: Chemical Comparison of Michipicoten Island Formation Samples to Experimentally Produced Liquids ................................ ................................ ................................ ..................... 81 x Figure 20: Liquid Extraction Diagram ................................ ................................ ........................ 82 Figure 21: Primitive Mantle Normalized Comparison Diagrams ................................ ................. 8 3 Figure 22: Ternary Magma Mixing Model ................................ ................................ .................. 8 5 xi KEY TO ABBREVIATIONS CC: Continental c rust DM: Depleted m antle DMM: Depleted MORB m antle HFSE: High field strength elements HREE: Heavy rare earth elements ICP - MS: I nductively coupled plasma mass spectrometer LA - ICP - MS: Laser ablati on - inductively coupled plasma mass spectrometer LILE: Large - ion lithophile element LIP: Large i gneous p rovinces LREE: Light rare earth elements Ma: mega - annums MCR: Midcontinent R ift MORB: M id - ocean ridge basalt REE: Rare earth element PPM: P arts per mi llion PM: Primitive Mantle SE: S tandard error SCLM: Sub - continental lithospheric mantle XRF: X - r ay f luorescence s pectrometer 1 Introduction Fundamental to our understanding of the plate tectonic model is that continents must break apart to f or m oceanic b asins. The process by which continents tear apart is known as continental rifting. Normal plate tectonic forces alone (i.e. , the tensile forces in plates caused by plate motion) are often insufficient to break the hard continental crust (Buck and Karner, 2004) . Thus , additional factors must be at play to generate a successful continental rift. One such factor is the impingement of a deep - seated thermo - chemical anomaly (i.e. , a mantle plume). These plumes allow for the localization of strain a nd heating of the continental crust making the crust the crust weaker and more susceptible to rifting (Buck and Karner, 2004) . The surface expression s of mantle plum e s are of l arge i gneous p rovinces ( LIP). Large i gneous p rovinces are defined as large ( >0.1 Mkm 2 ) , int ra p late magmatic provinces that are typically emplaced in a s h ort period of time (1 - 5 million years (henceforth Ma , mega - annums) ) , but with a maximum duration of < 50 Ma (Bryan and Ernst, 2008; Ernst, 2014) . The time between the end of volcanism and the start of rift opening can be quite variable, anywhere f ro m 2 Ma to 13 Ma (Courtillot et al., 1999) . T he hi ghly correlat ed timing between r ifting and the generation of large igneous provinces is a n indicator that the two processes are link ed . Stronger evi de nce that rifting an d LIP g en eration are l inked is the formation of triple junctions . Rift triple junctions are where three arms of a rift intersect , each arm being an area of thinned lithosphere due to gra b en or half graben formation . F or most continental rifts , the triple j unction s are both temporally and spatially link ed to the interpreted plume head center , usually within 1000 k m based on domal uplift (White and McKenzie, 1989; Courtillot et al., 1999) . This information makes for a compelling mod el for the link between rifting, plume head emplacement , and the generation of LIPs. The coupled plume - rift model for cont inental breakup requires a transition from active , plume - driven rifting to passive , plate - driven extension. This also requires a transition from fault - dominated strain accommodation to diking strain accommodation. This transition is typically 2 marked by a r enewed pulse of magmatic activity (Ernst, 2014) . Important questions remain about the nature of this magmatism , and what sources are contribut ing to said magmatism during this transitional period. Unfortunately, it is difficult to obtain samples of rock that record rift processes because with successful continental rifting , transitional volcanic material is submerged and buried beneath the ocean ic sediments (Stein et al., 2018) . While we can infer the presen ce of this magmatism th r ough the use of seismic tomography and magnetic anomaly mapping , sampling has proven difficult. Luckily there is a second type of rift known as a failed rift. These failed rifts do not succeed in forming oceanic basins and thus may preserve magmatism associated with the final stage of rifting in such a way that as to allow for sample collection. While rifts and their associated LIPs that succeed in forming oceanic basins are common, rifts that fail are far less numerous, but no less important . Of the known LIPs associated with failed rifts, some have been highly deformed and eroded , like the Emei shan traps (Courtillot et al. 1999). Others like the Circum Superior and Matache wan - Mistassini are eroded down to their plumbing systems. This leaves just the Keweenaw Large Igneous Provence , associated with the extension of the Midcontinent Rift (MCR) , and the Siberian Traps , which has been associated with the extension of West Sib erian B asin , as the only well exposed failed rifts on Earth . The Western Siberian Basin and the Midcontinent rift both have synrift volcanism but in the Western Siberian Basin th is volcanism is buried beneath intra rift sediments. In the Midcontinent Rift the synrift volcanics have been exposed do to uplift ing during later orogenesis (Saunders et al., 2005) . This makes the MCR the best target for sampling of late stage lavas in a failed rift systems. V olcanism associated with the MCR has been broken down into four primary stage s. First is the initiation stage , which has be en associated with plume head impact beneath the MCR . N ext is the early stage of volcan ism , which has been interpreted as the result of melting from a primary plum e source during initial extension (Cannon, 1992; Nicholson et al. , 1997; 3 Miller and Nicholson, 2013; Stein et al., 2015) . The next phase is know n as the hiatus phase which , as its name suggest , is marked by a lack of volcanism . This reduction in f l ux may be d ue to the reduced extension or large - scale magmatic underplat ing (Nicholson et al., 1997; Miller and Nicholso n, 2013) . After this period of relative quiescence , volcanism is renew ed with mixed plume depleted mantle source , which has been suggested by some authors as a n indicat ion of renewed extension within the rift (Cannon, 1992; Nicholson et al., 1997; Miller and Nicholson, 2013) . The final stage of volcanism is known as the late stage and has not had nearly the s ame level of rigorous study as the p re vious stages , despite its pote ntial to shed light on the aforementioned questions surrounding the final stages of rifting. The late stage volcanics are much more localized: exposures are only found in the Keweenaw Peni nsula (Lake Shore Traps) , the Schoder - Lusten basalts, and Michipicoten Island ( Michipicoten Island Formation ). Recent dating by Fairchild et al. (2017) has shown th at the youngest volcanic sequence in the Midcontent Rift is the Michipicoten Island Formation . The Michipicoten Island Formation is also the most significant volcanic sequence from the late stage of volc anism , and thus will be the focus of this study. Majo r and trace element s were analyzed to understand what pro cesses are affecting the magma within the crust so we can understand what is impacting the magma as it ascends to the surface. By understanding wh at crustal processes are modifying the magmas , we can better assess the potential mantle sources cont ributing to late stage volcanism . From previous work (i.e. Nich olson et al., 1997; Miller and Nicholson, 2013) we have an underst and ing that the source of magmatism is linked to the tectonic situation within the MCR, thus understanding the source will inform us about the what is happening tectonically within the MCR. 4 Geologic Background G.1 Geological S etting The Keweenaw Large Igneous Province erupted an estimated 2 million cubic kilometers of (Cannon, 1992; Stein et al., 2014) . Volcani c rocks were mostly deposited within the rift , f orming one of the most pronounced positive gravity anomalies in North America ( fig. 1 ). This volcanism occurre d over 30 million years from 1110 Ma to 108 0 Ma during a lull in the contemporaneous Grenville orogeny, making it an unusually long - lived large igneous province (LIP) (McLellan d et al., 2001; Rivers, 2008; Swanson - Hysell et al., 2019) . The MCR for its age of ~1.1 billion years, is exceptionally well preserved. Our understanding of the formation of LIPs typically comes from Permian or younger sills are left ex posed (Ernst 2014). This preservation provides a rare opportunity to examine a Precambrian LIP. G.2 Tectonic Setting The Midcont in e nt rift formed between the Shawinigan phase and the Ottawa p hase of the Grenville Orogeny, during a lull in orogenic activ ity. The pre - MCR crustal shortening associated with the Shawinigan phase ended at ~1140 Ma (Rivers, 1997) . However, t he beginning of the Ott awa phase of the Grenville Orogeny has proven difficult to precisely determine . Understanding when the Ottawa Phase began is crucial since crustal shortening during the Ottawa phase has often been invoked as the cause of MCR failure (Cannon, 1994; Ernst, 2014) . S ome authors (e.g. McLelland et al., 2001; Rivers et al., 2002) have suggested that the beginning of th e Ottawa phase began w ith the end of magmatism at ~1090 Ma but r ecent dating by Fairchild et al. ( 2017) has pushed back the timing o f the last phase of volcanism to 108 3 Ma. Thus, i ndependent constraints on when the Ottawa phase began , which is not dependent on simply the youngest magmatism , is required . 5 Granitic units closer to the Grenville front can give insight into when orogenesi s resumed in Laurentia. Dating of units from the Adirondack Highlands , such as the deformed Hawkeye Granitic suite , reveals ages between 1103 - 1093 Ma ; less deformed pyroxene bearing syenite has been dated at 1080 ± 4 Ma (Ch iarenzelli and McLelland, 1991) . The difference in the deformation is key to understanding when orogenesis must have begun. If the orogenesis had sta rted after the emplacement of the syenite, then both the granite and the syenite would have the same level of deformation. Since the pyroxene - bearing syenite has undergone less deformation , then it can be assumed that it formed after the Ottawa phase of or ogenesis had already begun (McLelland et al., 2001) . While dating of metamorphosed units from the Adirondacks indicate s when the Ottawa phase began, it does not nece ssarily indicate when rifting ceased. Swanson - Hysell et al. (2019) suggest that a series of angular unconformities, which has been roughly constrain ed to ~1091 Ma, represent the transition from active rifting to post rift thermal subsidence. If active rifting ceased at 1091 Ma , then why does magma tism continue to 1083 Ma on Michipicoten Island? G.3 Geologic E volution of the Midcontinent R ift Sever al divisions have been proposed to separate rift magmatism based on chemical composition (Shirey et al., 1994; Marshall, 1996; Davis and Green, 1997; Nicholson et al., 1997; Heaman et al., 2007; Vervoort et al., 2007; Miller and Nich olson, 2013) . Here we utilize the nomenclature proposed by Miller and Nicholson (2013) . This naming convention combines both geochemistry and geochronology to divide the magmatic activity into five di stinct groups: initiation, early, hiatus, main, and late stage. G.3.1 Initiation sta g e (1115 - 1110 Ma) The initiation of the Midcontinent Rift began at 1115 Ma with the emplacement of intrusions in the Nipigon Embayment area (Heaman et al., 2007) . Th e geochemical trace element 6 characteristics of these sills seem s to indicate that the plume impacted aro und this time (H ollings et al., 2007) . During this time period , only intrusions are preserved , wit h surface volcanics absent. It has been proposed that this may be due to an increase in the rate of erosion due to crustal doming (Mille r and Nicholson, 2013) . G.3.2 Early Stage (1110 - 1105 Ma) This group is represented by the first flows of Keweenaw LIP that are dominantly picritic in composition. The early flows from this group have relatively primi t ive compositions and Nd values n ear zero , which has been inferred to be the plume signature (Shirey et al., 1994) . However, up section , the composition of the lavas becomes progressively more negative in terms of Nd . These observation s have been interpr e ted as indicating a progressive increase in crustal contamination (Shirey et al., 1994) . It has been suggested that th i s indicates that the initial melts were able to move quickly through the cold continental crust, but as time progressed these basalts heated the crust and started assimilating it into the melt (Shirey et al., 1994) . Miller and Nicholson (2 0 13) interprets the isotopic data , along with elevated Th/Yb (i.e. >1 ), as the result of fractional crystallization and assimilation in deep crustal magma chambers. This is in agreement with existing models that suggest sialic magmas wer e generated as the result of deep crustal anataxis (Vervoort and Green, 1997; Vervoort et al., 2007) . G.3.3 Hiatus Stage (1105 - 1101 Ma) This time period is defined by the deposition of sediments and the absence of volcanism. A notable exception to this are the occurrence of Group 5 bas alts at Mamainse Point, and limited rhyolite volcanism throughout the rift (Shirey et al., 1994; Miller and Nicholson, 2013) . At Mamainse Point , where volcanism continued during this phase , much of the stratigraphic record is represented by sedimentary units . For example , the Great Conglomerate and the Basa ltic Clast Conglomerate occupies ~50% of the stratigraphy during this hiatus phase. Despite the 7 hiatus in volcanism , existing interpretation s sugges t that t he plume continued generating melt (Miller and Nicholson, 2013) . It has been suggested that during this time there was extensive ponding of mafic melts at the crust - mantle boundary , generating a ma gmatic underplate . This underplate can be observed geophysically in gravity models from the MCR region. G.3.4 Main Stage Volcanism (1101 - 1094 Ma) T he main stage represent s a time period of renewed and vigorous volcanic activity, e mplacing the majority of the volcanic material found within the MCR. Lava flows from the m ain s tage, like the first flows of the e arly s tage, show little evidence of crustal assimilation (Shirey et al., 1994; Nicholson et al. , 1997) . This has been interpreted as the result of long - lived crustal magma chambers having generated an insulating mar ginal zone, thus protecting the melt from contamination with in the continental crust (Miller and Nicholson, 2013) . As magmatism progr essed during this stage , magmatism became more primitive (e.g. became more Mg rich) (Paces, 1988; Paces and Bell, 1989a; Klewin and Berg, 1991) . This has bee n interpreted to be the result of an increase in magmatic flux and the plumbing system becoming better developed over time (Shirey et al., 1994; Miller and Nicholson, 2013) . Another observable trend see n throughout the rift during this time period is the change in isotopic values from near primitive Nd i values (plume - i near 0) towards more positive values ( deplete d upper mantle like). This change has been interpreted as the commencement of mixi ng between asthenospheric melt and melt from the plume (Paces, 1988; Paces and Bell, 1989b; Shirey et al., 1994) . G.3.5 Late Stage (1094 - 1080 Ma) Volcanism from this stage is localized as the Lake Shore Traps (LST) in th e Keweenaw Peninsula and the Porcupine Mountains in northern Michigan, and the Mich ipicoten Island Formation on Michipicoten Island Ontario (Fig: 2 ). In the Keweenaw and Porcupine Mo untains, 8 the LST overlies the main stage Portage Lake Volcanics and are in terc a lated with the Copper Harbor Conglomerate. On Michipicoten Island, the Michipi coten Island Formation overlies a series of late stage sills , which intruded the main stage volcani sm of the Quebec Mine Member . Miller and Nicholson (2013) interpreted these late stage volcanics to be the result of mix ing of remnant plume components with depleted asthenospheric components , t hough existing data supporting this argument is limited . Annells, (1974) concluded that the late stage lavas on Michipicoten Island were the result of basaltic melts mixing with remnant upper crustal melts based on major and minor elem ent data from a few flows, and a limited geochemical dataset. G.4 Geolog y of Michipicoten Island Michipicoten Island is the 3 rd largest island in Lake Sup erior , and it is almost entirely composed of igneous material associated with the Midcontinent Rift. Michipicoten Island is located on the most northern segment of the rif t eastern arm (Fig. 1 ) . The intra - rift volcanism dips towards the central rift axis and is bounded to the north by one of the rift graben faults known as the Michipicoten Island F ault. The igneous rocks of the island can be subdivided into three general fo rmations: the Quebec Mine Member , the Michipicoten Island Intrusives, and the Michi picoten Island Formation (fig. 3 ) . The Michipicoten Island Formation will be the focus of this study. The Mamainse Point Formation mostly consists of coarse - grained olivine tholeiite basalts flows ~855 m thick . Based on their similar characteristics, Mama inse Point Formation basalts have been thought to be stratigraphically related to the Mamainse Point Sequence (Annells, 1973) . This idea is further supported by paleomagnetic data . Fairchild et al. (2017) points out that the mean paleomagn eti c pole of the Mamainse Point Formation on Michipicoten Island (Palmer and Davis, 1987) overlaps with the mean pole from Mamainse Point Sequence (Swanson - Hysell et al., 2014) . The Michipicoten Island Intrus ives intrude the Mamainse Point Formation and were was empla ced at 1086.5 +1.3/ - 3.0 Ma (Palmer and Davis, 1987) , and make s up ~50% of the surface 9 area of the island (Annells, 1974) . The intrusive s were emplaced in two stages: the F irst I ntrusive P hase comprises a quartz porphyry and a felsite . T he S econd I nt rusive P hase is composed of a granitic phase and basaltic andesite phase. The Michipicoten Island Formation lies unconformably atop the intrusions. The Mamainse Point Formation is locally separated from the intrusions by a polymictic conglomerate. The Mic hipicoten Island Formation is composed of 5 major units: the Cuesta Member, the Channel Lake Member, the Queb ec Harbor Member, the South Shore Member, and the Davieaux Island Member, lying stratigraphically in that order (fig. 3 ). The Cuesta Member is a pl agioclase - phyric andesite that is divided into two flows referred to as the Cuesta U pper Flow and Cuesta L ower F low. The Cuesta Member is between ~255 m and ~ 340 m thick with the Cuesta Lower F low making up ~50% of the outcrop on the west side of the islan d but pinching out on the east side (fig. 3 ) (Annells, 1974) . The Channel Lake Member is comprised of aphyric basalt ic andesites ~260 m thick (Annells, 1974) . Individual flow s from this unit are between <1 m to 10 m thick with the exact number of flows uncertain. The next major unit is the Quebec Harbor Member a ~ 275 m thick aphyric andesite (Annells, 1974) . This member is comprised of multiple flows but due to poor preservati on and a lack of outcrops the number and thickness of flo ws cannot be determined . On the west side of the island between the Channel Lake Member and the Quebec Harbor Member is a small outcrop of lithic tuff that has been dated to be 1084.35 ± 0 .20 Ma (Fairchild et al., 2017) . Atop the Quebec Harbor Member is the well - preserved South Shore Member, an olivine - free and plag ioclase - phy ric basalt. The South Shore Member is ~ 265 m thick in tot al , this unit is comprised of twenty one individual flows that are <1 m to >30 m thick (Fairchild et al., 2017) . The youngest un it is the Davieaux Island Member , which forms a string of islands off the southern coast of Michipicoten I s land . The Davieaux Island Member is a single ~ 195 m thick flow of feldspar and quartz - rich r hyolite (Annells, 1974) . The exact relation between the S o uth Shore Member and the Davieaux Island 10 Member is unclea r since there is no contact between the two units. This rhyolite unit has been dated at 1083.52 ± 0 .23 Ma , making it the youngest flow in the MCR (Fairchild et al., 2017) . 11 Methods M.1 Major and Trace Element A nalyses A total of 78 samples were collected from Michipicoten Island. Samples were taken from flow interiors , when possible , to minimize the effect s of contamination from v esicle fill of secondary quartz and chlorite . Units sampled include the Upper and Lower Cuesta M ember, the Channel Lake Member , the Quebec Harbor Member, the South Shore Member and th e Davieaux Islan d Member. Samples were cut in to ~30g billets, then polis hed to remove the saw marks. Once saw marks we re removed , samples we re washed twice in deionized water in an u ltra s onic bath to remove any surface contamination. Samples were then crushed using a ste el jaw crusher , and subsequently powdered with an alumina mill. These powders were fused into glass disks using a lithium tetraborate flux at 1:3 sample: flux ratio followi ng the methods of Rooney et al. (2012) Major and trace elements concentrations were analyzed at Michigan State University . Major elements wer e determined using a Bruker S4 Pioneer X - Ray Fluorescence Spectrometer (XRF). These disks were then analyzed for t race elements using a Photon - Machines Analyte G2 Excimer laser and Thermo Scientific ICAP Q quadrupole inductively coupled plasma mass spectr ometer (ICP - MS). Standard d eviation on replicated analyses is less than 5% , except for low concentrations (< 2 pmm) of Cr and Ni. S ample s were run over four sessions. Major element s from the XRF were used as internal standard s for trace element analysis an d processed using Thermo Qtegra software. Machine drift was handled by analyzing the sample in triplicate and a ppl ying drift correction using known concentrations in geological standards BHVO - 2 and J B1a. M.2 Isotopic A nalyses Isotopic analyses were condu cted at the University of Hawaii at Manoa School o f Earth Science and Technology (SOEST) laboratory. Samples were first processed at Michigan State University 12 using a metal - minimal technique . Sample were cut into ~10g billets and crushed , then picked for f resh ness to reduce the effects of secondary alteration . Groundmass that was accepted for analyses was generally da rk and glassy without the presence of secondary mineral such as chlorite . The picked sample groundmass was then leached in 6M hydrochloric aci d for >16 hours at University of Hawaii at Manoa following a protocol modified from Koppers et al. (2003). Single dissolutions were used to sep a rate Pb Sr Nd isotopes following Konter and Storm (2014 ) and Hf isotopes following Connelly et al. (2006). The a nalyses took place on a Nu Plasma HR multi - collector inductively coupled plasma mass spectrometer (MC - ICPMS) . The initial separation of Pb and Sr from the rest of the matrix took place us ing a Sr - Sp ec resin c olumn. The cut that was not used for the Sr - Pb cleanup would be later used for the Nd and Hf separation. The Sr cut went th r ough a second column of Sr - Spec resin to remove any remaining interfering elements , following the same procedure as the fi rst separation but without the lead step. Sr isotopic ana lyses w ere normalized to 86 Sr/ 88 Sr ratio of 0.1194 to correct of mass fractionation . Kr interferen ce was correct ed for using a multi - dynamic program developed by (Konter and Storm, 2014) . The blank on Sr analysis was ~70 pg. After initial Sr - Pb separation , the Pb cut is pa ssed though AG1 - X8 resin to separate any remaining unwanted elements. Tl spike (NIST SRM 997; 205 Tl/ 203 Tl = 2.3889) was used to correct for drift. The blank on Pb analysis was ~70 pg. After the Sr and Pb step s, Nd was the next element to be r emoved from t he matrix. This matrix is dissolved in ascorbic acid to reduce the Fe 3+ to Fe 2+ , then passed through two columns of TRU resin , which separate s and purifies the REE from the matrix. To separate the Nd from the REE , the REE cut was placed in a c olumn of LN r esin for chromatographic separation. The analysis of Nd used a multi - dynamic program to correct for REE interference and mass bias , normalizing samples to the accepted present - day 146 Nd/ 144 Nd ratio of 0.7219 (Konter and Storm, 2014) . Drift was correct ed for using Jndi - 1 sta ndard using standard sample bracketing. Blank on Nd was ~ 40 pg. 13 The remaining sample left over after the previous separations was used for the h afnium separation f o llowing a modified method from Connelly et al. (2006) ; the full method is described in Finlayson et al. (2018) . Each s ample is placed on a bed of AG50 - x8 resin and washed with a HC l - H F mixture to remove the high field strength elements ( HFSE ) from the rest of the matrix. The final treatmen t is to remove the Ti from the sample , because refractory Ti can build up on the sample cone and creat e an electrical barrier to Hf , generating im precise Hf results (Blichert - Toft et al., 1997) . To remove the Ti , the HFSE cut is place d on a bed of DGA resin and Ti is removed by wash in g sample with 3.5 M HNO 3 . M.3 Isotopic A ge C orrection While the act of correcting isotopic data is commonpla ce in the geochemistry community , m ost studies their corrected age . T his is compounded by the fac t that many of the published equations assume that a researcher measured the parent / daughter ratio using isoto pe dilution . O d ilutions. The lack of detailed method on how a researcher conducted their age correction can lead to a reproducibility problem. Thus, I outline in detail the methods used for age - correcting our data . Like all a ge corrections for an isotopic system , we assume that the isotopic systems has remanded closed since eruption . The equation for obtaining the ori ginal radiogenic isotopic ratio is: (equation 1.1) i is the original daughter isotopic ratio and I is the daughter isotopic ratio today. R is the parent isotopic ratio. 14 M.3.1 Rb - Sr System We do not measu re the parent isotopic 87 Rb/ 86 Sr on the Multi - Collector ICP - MS , so it has to be derived from the 87 Sr/ 86 Sr ratio and the Rb and Sr concentration (Faure, 1986) : (equation 2 . 1 ) Where C refers to concentration and wt refers to atomic weight (de Laeter et al., 2003) . Both the atomic weight of Sr and atomic abundance of 86 Sr are dependent on the 87 Sr/ 86 Sr o f the samples. First arranging all Sr isotopes in terms of the most abundant isotope i. e. 88 Sr. For the stable isotopes of Sr we used the natural abundances (de Laeter et al., 2003) . We can use the natural abundance because , while their true abundance will change based on the 87 Sr/ 86 Sr the ratio , stable isotopes will remain constant . For the radiogenically produced 87 Sr the 87 Sr/ 88 Sr ratio can be determined using the measured value: (equation 2 . 2 ) T he abundance of the different Sr isotopes in th e sample can be determined by: 15 (equation 2 . 3 ) Where This allows us to determine the atomic weight of Sr in this sample , as required for equation 2 . 1 : (equation 2.4 ) T he atomic weight of an ele ment is equal to the sum of the mass es (m) of each isotope multiplied by the abundance of the isotope . All terms are dependent on the 87 Sr/ 86 Sr ratio and are required for equation 2 . 1 have been solved for . M.3.2 Sm - Nd System We do not measure parent i sotopic 147 Sm/ 144 Nd on the Multi - Collector ICP - MS so it has to be derived from the 143 Nd/ 144 Nd ratio and the Sm, Nd concentrations. 16 (equation 3 . 1 ) Where C refers to concentration and wt refers to atomic weight (de Laeter et al., 2003) . Both the atomic weight of Nd and atomic abundance of 144 Nd are dependent on the 143 Nd/ 1 44 Nd of the samples. First expressing all Nd isotopes in terms of the most abundant isotope i.e. 142 Nd. For the nonradiogenic isotopes of Nd we used the natural abundance s. We can use the natural abundance for this because , while their true abundance wi ll change based on 147 Nd/ 144 Nd , the ratio between stable isotopes will remain constant (de Laeter et al., 2003) . For the radiogenically produced 143 Nd isotope the 143 Nd/ 142 Nd ratio can b e determined using the measured 143 Nd/ 144 Nd (equation 3 . 2 ) The abundance of the different Nd isotopes in this sample can be determined by: 17 (equation 3 . 3 ) Where This allows us to generate the atomic weight of Nd in this sample required for equation 3 . 1 (equation 3 . 4 ) The ato mic weight of an element is equal to the sum of the masses (m) of each isotope multiplied by the abundance of the isotope . All terms are dependent on the 143 Nd/ 144 Nd ratio and are required for equation 3 . 1 have been solved for. 18 While the initial 143 Nd/ 1 4 4 Nd is an important ratio , the limited fractionation between Nd and Sm coupled with the long half life of 147 Sm , make s the difference between the 143 Nd/ 144 Nd i of individual samples quit e small . Thus, commonly 143 Nd/ 144 Nd ratio is expressed relative to the chondrite uniform reservoir (CHUR) , which a s the name suggests , is defined as a mantle source with the same Sm - Nd concentrations and isotopic ratio s as C1 chondrite (Bouvier et al., 2008) . When comparing the 143 Nd/ 144 Nd of the sample to that of CHUR , differences between samples b ecome larger. This comparison is referred to as epsilon notation : To use this notation properly , the samples must be corrected to initial 143 Nd/ 144 Nd , compare the initial 143 N d/ 144 Nd of our sample to the CHUR. So, we must age correct 143 Nd/ 144 Nd CHUR as well. T he 147 Sm/ 144 Nd of CHUR has already been defined by Bouvier et al. (2008) and thus does not need to be derived. M.3.3 Lu - Hf System We d o not measure parent isotopic ratio 1 76 Lu / 1 77 Hf on the Multi - Collector ICP - MS so it has to be derived from the 1 76 Hf / 1 77 Hf concentration of Lu and Hf . (equation 4 . 1 ) 19 Where C refers to concentration and wt refers to atomic weight (de Laeter et al., 2003) . Both the atomic weight of Hf and atomic abundance of 177 Hf is dependent on the 176 Hf/ 177 Hf of the samples. First expressing all Hf isotopes in terms of the most abundant i s otope i.e. 180 Hf. For the stable isotopes of Hf we used the natural abundances (de Laeter et al., 2003) . For the radiogenically produced 176 Hf isotope the 176 Hf/ 177 Hf ratio can be determined using the measured value. (equation 4.2 ) The abundance of the different H f isotopes in this sample c a n be determined by: (equation 4.3 ) 20 Where This allows us to generate the a tomic weight of Hf in this sample required for equation 4 . 1 (equation 4.4 ) The atomic weight of an element is equal to the sum of the masses (m) of each i s otope multiplied by the abundance of the isotope . Now we have derived all the terms required for equation 4 .2 to be solved , which in t u rn allows for 4 .1 to be completed. Like th e Sm - Nd system the Lu - Hf system , has a long half - life and little fractionati o n , thus it is commonly put into epsilon notation. (equation 4.5 ) 21 Using the equation 1 .1 the CHUR 176 Hf/ 177 Hf can be corrected for the age of the sample. We used 176 Hf/ 177 Hf and 176 Lu/ 177 Hf for CHUR from Bouvier et al. (2008) . M.3.4 U - Th - Pb System Because we measured all of the Pb isotopes and 204 Pb is the only non - radiogenically produced isotope of Pb , we can use a different method th a n the one used for the other isotopic systems . (equation 5 .1 ) Since the abundance of the radiogenic isotopes are solved , the abundance of 204 Pb is: Now that we know the abundances of the different Pb isotopes are we can find the atomic weight o f Pb in our sample 22 (equation 5.2 ) The atomic weight of an element is equal to the sum of the masses (m) of each isoto pe multiplied by the abundance of the isotope . As with the previous isotopic systems we need to find the parent stable isotope ratio (equation 5.3 ) The equation for the 235 U/ 204 Pb can be simplified because 238 U/ 204 Pb was calculated. M.4 Isotopic Mixing Most continental magmatism is not derived from solely one source , but are rather a mix between two or more source s . Thus, to model the interaction between source s we need equations that describe what the mixing lines should look like , which we can then compare to our data. 143 Nd/ 144 Nd and 87 Sr/ 86 Sr will be used as an example , but this can be done for any of the common heavy radiogenic isotopic systems. If we start with mixing between t wo reservoirs for 87 Sr/ 86 Sr . (equation 6 .1) 23 Where: is fraction of component b Equation 6 .1 simplified if we assume that 86 Sr is equivalent to Sr conce n t r ation ( 86 Sr ~ Sr) (equation 6 .2) Appling this to the Nd system (equation 6 .3) Equations 6 .2 and 6 .3 can be used to solve for the mixing iteratively . To de fine the equa tion that describes the mixing line between endmember s we must combine equation 6 .2 and 6 .3 into a single equation. If we rewrite equation 6 .2 to solve for f b We can then place it into equation 6 .3 to have one uniform equation that describes the mixing line between th e two components. (equation 6 .4) Which can be simplified to 24 25 Results R.1 Petrography Here , we use the organization scheme originally proposed by Annells (1974) , which divides the different units on Michipicoten Island largely based on petrological an d mineralogical differences. The order of this section f o llows the order of eruption of the Michipicoten Island Formation . The Cuesta Lower Flow has an abundance of large glomerocr ysts and phenocrysts, which make up about 50% of the volume of the sample (f ig. 4 ) . Both p henocryst and glomerocryst phases include plagioclase, clinopyroxene, orthopyroxene, biotite and oxides. Plagioclase phenocrysts are euhedral and oscillatory zoned; s ome exhibit sieve texture with occasional clinopyroxene and/or oxide chadocr yst s . Both clinopyroxene and orthopyroxene phenocrysts phases are anhedral. Oxides and biotite phenocryst phases are subhedral. Groundmass within the Cuesta Lower Flow is primarily plagioclase and oxides . Like the Cuesta Lower Flow, the Cuesta Upper Flow h as an abundance of large glomerocrysts and phenocryst phases, but phenocrysts only make up about 40% of the sample by volume. Both p henocryst and glomerocryst phases include plagioclase, clinopyroxene, oxides , and what appear to be replaced olivine crystal s (fig. 4 ) . These sparsely replaced olivine crystals seem present in both the glomerocrysts and as phenocryst phases but can be difficult to distinguish from vesicles because the vesicles are filled with a similar alteration product. The plagioclase phenoc rysts are euhedral, oscillatory zoned, and some are sieved with occasional clinopyroxene and or oxides forming in the spaces. Clinopyroxene phenocrysts are subhedral and the oxide phenocrysts are anhedral. Groundmass in the Upper Cuesta Member is comprised primarily of plagioclase and oxides. The Channel Lake Member is largely aphanitic, with phenocrysts making up <1 - 5% of the volume of the samples (fig. 4 ) . Phenocryst phase s include plagioclase, clinopyroxene and oxides. Plagioclase phenocrysts are euhed ral , some with sieve or skeletal texture and 26 sometimes contain ing clinopyroxene grains. Both clinopyroxene and oxide phenocrysts are anhedral and the oxides are often embayed . The g roundmass contains plagioclase, oxides and minor clinopyroxene in the lower flows; clinopyroxene is absent in the upper flows. S parsely distributed coarser grained semi - spheric al mesostasis is present within some of the flows. This mesostasis has ir regular shapes and boundaries, containing plagioclase and minor clinopyroxene. Som e of the clinopyroxene that comprises the mesostasis is bladed in texture. The West Sand Bay Tuff member lies between the Channel Lake and the Quebec Harbor Members. This tu ff is comprised of subangular green and red clast s with a light green matrix. This matrix is fined grained with minor plagioclase and quartz. The clasts have been highly altered but appear to be vesicular basalt and quartz porphyry. Annells (1974) reports identifying clasts from the Cuesta flo ws , but this is not apparent in our sample s . The lower flows of the Quebec Harbor Member are com paratively phenocryst rich, up to 10% phenocrysts (fig. 4 ) . Two samples were taken from the lower flows (fig . 3 ). Phenocryst phases in the lower section consis t of plagioclase, clinopyroxene and oxides. The plagioclase crystals are often euhedral, oscillat ory zoned , and with sieve texture. While most are euhedral, some are anhedral with highly rounded edges. Clinopyroxene in the lower flows are anhedral, embayed and occasionally elongated. Oxides are euhedral and occasionally embayed. Cumulates, mostly cons isting of plagioclase with minor oxides, are common. Annells (1974) present in our samples. Groundmass phases inclu de plagioclase, clinopyroxene and oxides. The upper flows of the Quebec Harbor Member are phenocr yst poor (<1% volume crystals) and are often altered. The result of this alteration is a groundmass made up of oxidized hematite and chlorite. Where phenocrys ts are present, they are euhedral plagioclase. The groundmass consists of plagioclase, oxides and minor clinopyroxene. The South Shore Member has few phenocrysts, only making up ~1% of the sample volume (fig. 4 ) . The phenocrysts consist of plagioclase, clinopyroxene and occasional oxides. 27 Most of the phenocrysts are isolated crystals, but there are so me cumulates. Plagioclase and oxide phenocrysts are subhedral to anhedral, while clinopyroxene is anhedral. Groundmass consists mostly of p lagioclase, clinopyroxene , and oxides. T he groundmass commonly shows plagioclase in a trachytic texture . The Davieau x Rhyolite is microcrystalline, with no phenocryst phases that we have observed (fig. 4 ). Layers of oxides and silicates are folded and sho w plasticity of the rhyolite hat appear to be secondary based on oxidation rims around these lenses. Microcrystalline phases include quartz, potassium feldspars and alt ered oxides. Annells (1974) also observed the texture of the Davieaux Rhyolite and did report rare highly altered fel dspars phenocrysts . Annells (1974) indicated that the lack of well - preserved glass in this unit suggests it has been subjected to devitrifi cation from hydrothermal activity . R.2 Geochemistry While the groupings of the Michipicoten Island Formation were made on the basis of petrography, these divisions are also apparent in the geochemistry. Rock type classification of samples is based on Le Bas et al. (1986) . The Michipicoten Island Formatio n follow s a subalkaline tholeiitic trend (fig. 5 ) R.2.1 Major Elements The oldest major volcanic unit in the Michipicoten Island Formation is the Cuesta Mem ber , which is divided into the Upper and Lower F low. The lower flow is andesitic in composition based on total alkali - silica diagram, with SiO 2 conce n t r ation of ~57 wt. % and 5.5 wt. % K 2 O wt. % + Na 2 O wt. % (fig. 5 ) . Although samples were taken at differe nt stratigraphic positions , it appears , based on the lack of signif icant geochemical changes in the major or trace elements between samples , that these samples are all from the same flow. The Upper Cuesta F low is 28 also an andesite, but with a higher SiO 2 co ntent of 61 wt. % (fig. 5 ) . In major element variation diagrams the Questa Lower F low lies between the South Shore Member and the Channel Lake Member , or with the more mafic Channel Lake samples. T he Questa Upper flow plots with the Channel Lake Member , bu t with the slightly more s ilica rich compositions (fig. 6 ) . The next youngest unit is the Channel Lake Member. The unit is chemically the most variable of any unit in the Michipicoten Island Formation in total alkali - silica space, ranging from low - SiO 2 and esites to rhyolite (SiO 2 wt. % 57 - 74 avg. 64.5 wt. % ) (fig 5 ). The most evolved flow occurs near the bottom of the section , while the least evolved flows occur near the top of the sequence. The Channel Lake Member has a well - developed trend line with the n ormally compatible elements MgO, FeO, TiO 2 , and CaO all decreasing with SiO 2 . P 2 O 5 also exhibits a decreasing trend (fig. 6 ) . The normal ly incompatible elements K 2 O and Na 2 O increase with silica. The lower SiO 2 wt. % samples have an increasing trend in Al 2 O 3 wt. % when compared to SiO 2 wt. %, but the samples with more than 60 wt. % SiO 2 have a decreasing Al 2 O 3 trend with increasing SiO 2 . The flow - by - flow diagram ( f i g . 7 ) shows a general decrease in SiO 2 concentration up section , but faulting of this par ti cular unit is common and may complic ate any interpretation (Annells , 1974) . Above the Channel Lake Member is the Quebec Harbor Member. Only a few flows of the Quebec Harbor Member were collected due to limited accessible exposure. This unit is dacitic in composition (fig. 5 ) . This u nit has been divided into an unaltered lower section and an altered upper section. Two sample s were taken from the unaltered lower part, and compositionally they appear to be a part of the same flow . Samples from this unit are dacitic in composition (~ 7 0 wt. % SiO 2 ), breaking the general trend seen in the Channel Lake Member of becoming less evolved up stratigraphy (fig. 7 ). In major element space the Quebec Harbor Member samples plot along with the Channel Lake Member dacites (fig. 6 ) . The South Sho re Me mb er is the most mafic of any of the units in the Michipicoten Island Formation , ranging from basaltic to basaltic andesite in composition (2.1 - 5.5 wt. % MgO, 49 - 56 29 wt. % SiO 2 ). This member becomes slightly more SiO 2 and TiO 2 rich up section. A sawtooth p at tern is evident in the flow by flow diagrams of decreasing incompatible elements followed by a sharp increase; this is the opposite for the compatible elements. This pattern can be most easily seen in SiO 2 and CaO (fig . 7 ). Like t he Channel Lake Member, th e South Shore M ember has a clear trend line , and the two trend lines intersect and about ~56 wt. % SiO 2 to fr o m a nearly continu ous Michipicoten Island Formation trend. In most major element spaces, the Channel Lake Member and the South S h ore Member beha ve similarly and with identical slopes . MgO, FeO, and CaO decrease , while K 2 O and Na 2 O increase with SiO 2 . Unlike the Channel Lake member , TiO 2 does not change when compared to SiO 2 and the major ity of South Shore Member Samples f orm an increasing trend in P 2 O 5 v s . SiO 2 (fig. 6 ) The Davieaux Island Member is a r hyolite and is the most silica rich (72 wt. % SiO 2 ) of any of the Michipicoten Island Formation units , except the samples from the Channel L ake M ember , which appear to be altered (fig. 5 ) . Unlike the r est of the Michipicoten Island Formation the Davieaux I sland rhyolite is alkaline instead of sub - alkaline. The rhyolite lies at the end of many of the major element trend lines like FeO, TiO 2 and Al 2 O 3 , but falls off the trend line in Na 2 O, CaO and K 2 O ( fi g. 6 ) . R.2.2 Trace elements The Cuesta Upper and Lower flows of the Michipicoten Island Formation exhibit moderate concentration s in the first row transition elements and in high field strength element s (HFSE) , when compared to the other Michipicote n Island Formation samples in most trace element spaces ; and often overlap with the more mafic samples of the of the Channel L ake M ember (fig. 9 , 10 ) . Thi s is not true , however , for Ni and Cr in the Cuesta Lower Flow , which has higher concentration s of the se elements in comparison to the majority of the other units. Both the Questa Upper and Lower Flows have higher Eu concentration s th a n the other Michipicoten Island Formation units. The samples taken from the u pper part of the Cuesta U pper F low are 30 quite v ar iable in the l arge i on l ithophile elements (LILE) ( fig. 8 ) . From the primitive mantle normalized trace element diagrams we can see that the Cuesta members have strong negative Ti and Sr anomalies (fig. 11 ) . T he chondrite normalized REE diagram shows tha t the flows have a high Eu concentration when compared to other samples , with similar concentrations in REE (fig. 12 ) . The Channel Lake Member, which lies above the Cuesta Flows, is the most chemically divers e unit of any from the Michipicoten Island For ma tion . From the array formed by the Channel Lake member , a trend line can be observed in both the first row transition elements ( decreasing with increasing SiO 2 ) and HFSE ( increasing with increasing SiO 2 ), except for Eu , which decreases with increasing Si O 2 (fig. 9 , HSFx) . The Channel Lake trace element concentration s often lie between the South Shore Member (most mafic) and the Davieaux Island Member (most silicic) , with the exception of some HFSE where the Channel Lake Dacites have the highest concentrat io n of Zr, Hf, U, a nd LREE. Primitive Mantle normalized values appear similar to those of the Cu esta Member, but most have stronger Sr and Ti anomalies (fig. 11 ) . I n the Chondrite normalized REE diagrams, the Channel Lake Member has a greater Eu anomaly (f ig . 12 ) . The Quebec Harbor Member is more restricted in its trace element concentrations th a n the Channel Lake Member stratigraphically below it. The unaltered samples from the Quebec Harbor M ember have relatively low first row transition element concent ra tions but hig h HFSE concentrations (fig. 9 , HFSEx). The unaltered Quebec Harbor Member trace element patterns overlap with those of the Channel Lake Member d acites. The trace element p atterns in the Primitive Mantle and Chondrite diagrams are comparable to those of the Channel Lake d acites (fig. 11 and 12 ). Samples from the altered section show how pervasive the alteration can be , causing samples to fall o f f the trend line even in the f luid immobile elements like Zr and Hf (fig. HFSEx ) . It should be noted t hat sample TOR0000QL , despite being taken from the altered region as denoted in Annells (1974) of the Quebec Harbor Member , mostly plots with un altered 31 samples in HFSE and first order transition element space but falls o f f the trend line in large ion lithophiles (fig. 9 , 10 , 8 ). Lying atop the altered sec ti on of the Quebec Harbor Member is the South Shore Member. This unit is the next most variable geochemically after the Channel La k e Member. This unit also forms a trend line in trace element space like that of the Channel Lake Member , with the same trajec to ry. The South Shore Member has the highest Sc, V and Co and these elements decrease with SiO 2 , but very low Ni and Cr , which var ies little with SiO 2 (fig. 9 ) . The HFSE in the S outh Shore Member are the lowest of any Michipicoten Island Formation unit and i ncrease in concentration with SiO 2 (fig. HSFx) . This unit has the weakest Sr and Ti anomalies on the Primitive Mantle diagram ( fig. 11 ) and only a slight Eu anomaly in the Chon drite normalized REE plot (fig 11 , RIx ) The Davieaux Island Member forms the to p unit of the Michipicoten Island Formation stratigraphic section and is geochemically distinct. This unit has a low concentration of the first row transition elements , often forming the most extreme samples of the trend line created by the other Michipi co ten samples (fig. 9 ) . The Davieaux Island Member HFSE concentrations are often off the trend line above ( i.e. Y, Th, Ta) or below ( i.e. Zr, Hf, U, La, Eu) but also lie at the end of the trend line in some spaces (Nb, Yb) (HSFx) . From the Primitive Man tl e normalized trace element diagram, we can see the Davieaux Island Member has the strongest Sr and Ti anomaly of any of the Michipicoten samples (fig. 11 ). From the chondrite normalized REE diagram we can observe that the Davieaux Island Rhyolite is depl et ed in LREE when compared to the Channel Lake dacites but has enriched HRE E as well as the strongest Eu anomaly (fig. 12 ). R.3 Isotope geochemistry Eight samples were chosen for analysis of Nd, Sr, Hf, and Pb isotopes. One sample was chosen from each o f the thinner units of the Davieaux Island Member, Quebec Harbor Member, 32 Cuesta Upper and Lower Members. From the larger units (i.e. the Channel L ake M ember and the South S hore M ember ), two samples were chosen. Isotopic data is age corrected to 1100 Ma. S a m ples from the Michipicoten Island Formation have radiogenic ( 143 Nd/ 144 Nd ) i ratios, except for sample TOR0000S1. This same trend can be seen in the Hf isotopic data (fig. 13 ). The Cuesta, South Shore, Davieaux Island Members lie s above the mantle array in Nd i i space, while the sample from the Quebec Harbor Memb er lie below. The Channel Lake Member has one sample above the mantle array and one just below (fig 13 ). Most of the samples have near or greater th a n chondritic i ( - 0.5 - +6.5) i ( - 1.5 - +4.0) values for sample TOR0000S1, which has negative i ( - i ( - 12.2) . This is unexpected i i of any sample reported from the MCR. Most o f the Michipicoten Island Samples have near chondritic (0.703125) 87 Sr/ 86 Sr initial values , except for TOR0000S1 from the South Shore Member and the sample from the Davieaux Island Member (fig. 14 ) . TOR0000S1 has a more radiogenic ( 87 Sr/ 86 Sr) i and plots c loser to rocks from Wawa Greenstone Belt adakites, andesites and basalts in the ( 87 Sr/ 86 Sr) i (Ture k et al., 1982; Polat and Münker, 2004) . The sample from the Davieaux Island Member has a less radioge nic ( 87 Sr/ 86 Sr) i value and plots away for the other sample (fig. 14 ) . Age - corrected Pb isotopic data ( 207 Pb/ 204 Pb vs. 206 Pb/ 204 Pb) i shows that samples from the Michipicoten Island Formation vary over a large range and lie above the northern hemisphere ref erence line (NHRL) (fig. 15 ) . In comparison to the Mamainse Point data, the Michipicoten Island samples lie in a parallel array to t he Group 5 samples from Mamainse point. Group 5 is the most crustally - contaminated series of lava flows in the Mamainse poin t sequence (Shirey et al., 1994) . Analyses of Mamainse P oint ( 207 Pb/ 204 Pb vs. 206 Pb/ 204 Pb ) i data shows that values radi ate from a common point ; the samples fr o m Michipicoten Island, however , do not radiate from this point . The Davieaux Island Member plots at the upper end of the Michipicoten 33 Island array in the ( 207 Pb/ 204 Pb vs. 206 Pb/ 204 Pb ) i space. In the ( 208 Pb/ 204 Pb vs. 206 Pb/ 204 Pb ) i space it plots well below the NHRL . The Cuesta Upper flow lies slightly above the line. The rest of the Michipicoten Island Formation sa mple s lie along the NHRL, and in a similar region as the Mamainse P oint samples (fig . 15 ) 34 Discussion D.1 Effects of Alteration D.1.1 Major and Trace Elements To understand the geochemistry of the Michipicoten Island Formation we must first assess what effects alteration has had on the different units. The presence of secondary alteration within the Michipi c oten Island Formation is well noted (Annells, 1974) . The m ajority of the alteration types seem to be chloritization and silicification. The first physical manifestation of alteration is found within the core of the Cuesta Upper Flow , where there is a secondary deposit of pyrite. Stratigraphically above this depo s it , the Cuesta Upper Flow becomes red an d plagioclase is the only remnant phase in this section. The other phases appe a r to have become red oxidized hematite and chlorite. Comparing the chemistry of a sample from the lower unaltered section (TOR0000QS) an d a sample from the upper altered section (TOR0000QQ) on a isocon diagram we can observe element mobility between samples (fig . 1 t1 ) (Grant, 2005) . From the isocon diagram it appears that Mg, Rb, K , Ba, Ca, Pb, Cs and Sr were mobil e based on thes e sample plotting away from the concentration of the a ltered sample (C a ) equal to the concentration of the unaltered sample ( C a = C m ) line. The alteration observed in the Upper Cuesta F low continues into the base of the Channel Lake Member. This has resulted in the unit being a similar red color . It also has c h alcedony and quartz being deposited in thin fractures and veins , as well as forming agates. At the top of the Channel Lake Member , alteration is also observed , changing the color of the basalt groundmass to dark reds and greens. Closely spaced veining o f calcedony, quartz and calcite are within these fractures . Within some of the flows , veining is so pervasive that the sample appears brecciated. Based on the isocon diagram , Pb, Ca, Sr, K, Rb were mobile (fig. 16 ) . The Quebec Harbor Member has the larges t amount of alteration of any of the units within the Michipicoten Island Formation . The physical result of t his alteration is that much of 35 this unit is a bright red color . This alteration allowed for preferential erosion and resulted in the formation of m a ny of the bays on the s outh s ide of Michipicoten Island (fig. 3 ). By c omparing the least altered of the alte red samples from the altered section of the Quebec Harbor Member and the unaltered Quebec Harbor Member samples , we can observe that K, Rb, Ca, Sr a nd Mn appear to have been mobilized ( fig. 16 ) . When we compare the most altered Quebec Harbor Member to the least altered, the elements fall completely off the C a = C m line . This may indicate th at alteration was so pervasive that it change d the mass of th e sample or that the samples are not comagmatic , thus changing the relative pro portions of the different elements and pulling all the samples off the C a = C m line. Above the altered section of the Quebec H arbor M ember, the South S hore Member is far less al t ered . Some alteration was concentrated around the edges of columnar jointing , with visible alteration penetrating about 5 mm into the basalt. Two of the flows were highly altered , appearing deep red. Based on the isocon diagram , K, Rb, Sr, P and Cs have b e en mobilized . Interestingly, Ni and Co fall off the C a = C m line ; since these elements are typically relatively immobile elements , this may indicate the difference in proportions of oxides between the two sample s . Only one sample was taken from the Daviea u x Island Member , which makes an isocon plot impossible. There are strong phys ical and chemical indicators that this unit has been altered with respect to its LILE . T his is evident in the strong positive Pb, K, and Rb anomal ie s in the primitive mantle diag r am. Fractures within the Davieaux Island Member are bleached white and kaolin ized, and some of these fractures are filled with quartz or calcite (Annells, 1974) . D.1.2 Effects of A lteration on R adiogenic I sotope T racers The effects of alteration are also present in the isotopic systems. The Lu - Hf and the Sm - Nd s ystems are based upon relatively fluid immobile elements , and thus are unlikely to have been impacted by the alteration processes. These elements are usually close to the isocon lines (fig 36 iscx). The Rb - Sr isotopic system and the U - Th - Pb isotopic system ha ve more potential for disruption due to alteration because of the relatively fluid mobile element s in the se system s . To help to assess whether the varying isotopic systems are open o r closed , isochron plots were utilized . These plots come with some assu mp tion s that must be kept in mind. One of the assumptions is that the individual samples are co magmatic (i.e. all coming from the same source ). The second assumption is that all the sample s are of the same age (or at least to within error of each other) (Faure, 1986) . Whole rock Rb - Sr isochron ages of the Channel Lake Member were conducted by Chaudhuri and Faure (1967) then later corrected by (Baragar, 1978) . T he data from th ese units were imprecise , resulting in the generation of an age that is clearly incorrect of 887 ± 78 M a ; more accu ra te zircon dating places the age of the Channel Lake Member at no young er than 1084.35± 0.20 Ma (Fairchild et al., 2017) . When w e apply a linear line of regression to our samples from the Michipicoten Island Formation , we get an age of ~ 854 Ma . Based on more accurate zircon dating we know actu al age of the Michipicoten Island Formation is somewhere between 1086 and 1083 Ma this wo ul d , indicate that the isotopic system has been perturbed in some fashion so as to reduce the apparent age from the w hole rock chemistry (Pa lmer and Davis, 1987; Fairchild et al., 2017) . A n explanation for calculating a young date might be that the samples are not all comagmatic , which is the first order assumption for gen erating a isochron age ; additionally any perturbation that resulted in a decrease in the Rb concertation or increase in Sr concentration post eruption would alter 87 Rb/ 86 Sr , and cause a lower ing of the slope and thus causing lower isochron age ( e q n . 2 . 1 ) . From the isochron d iagram s , the Rb - Sr system appears to be mostly clos ed , with most samples plotting near the chondritic line. Sample TOR0000S1 plots above this line , which may indicate some element mobility, or it may indi cate that the source of magmatism for this sample was more radiogenic (fig. 17 ) . The Davieaux Island Me mber plots below the chondritic isochron 37 line in an anomalously high 87 Rb/ 86 Sr vs. 87 Sr/ 86 closed for the Davieaux Island Member (Faure, 1986) . For both the Sm - Nd and Lu - Hf isotopic systems , the isochron diagrams show that the systems appear to be closed (fig. 17 ). The two So uth Shore Member samples fall the furthest off the chondritic isochron. As discussed above , falling o f f t necessarily the result of an open system but may be due to mixtures of non - chondritic reservoirs . With the South Shore Member, it would s eem that TOR0000S1 mixed with a low 147 Sm/ 144 Nd source resulting in low 143 Sm/ 144 Nd at present. This same pattern holds true for the Lu - Hf isotopic system , but TOR0000S1 falls even further from the chondrite isochron. The Pb isochrons show that the U - Th - Pb isotopic systems do not seem completely closed in some samples . The Cuesta Upper F low and Quebec Harbor Member plot away from the chondritic isochrons in all the Pb isotopic plots (fig. 17 ). Th is indicat es that the Pb system was open for the Cuesta Upper F low and Quebec Harbor Member . The Davieaux Island Member appears to be only perturbed in 238 Th/ 208 Pb v. 208 Pb/ 204 Pb (fig. 17 ). This would seem to indicate there was some thorium mobility in the r hyolite. This also explains why in the in 208 Pb/ 204 Pb v. 206 Pb/ 204 Pb space the Davieaux I sland Member falls far below the NHRL. In the other isotopic spaces, the Davi eaux I sland member plots where one would expect along the isochron based on where the other Michipicoten samples plot. From the isochron plots some samples isotopic signatures are unlikely to be solely the original magma composition . The 87 Sr/ 86 Sr and 2 08 Pb/ 204 Pb ratios from the Davieaux Island Member have likely been affected by some secondary process. It also appears likely that secondary prosses have affe cted the Cuesta Upper F low and Quebec Harbor Member Pb isotopic data. For the Davieaux Island Memb er, Cuesta Upper F low and Quebec Harbor Member the isotopic systems that appear to have been impacted by secondary prosses data are plotted but should not be used for interpretation. 38 D.2 Generation of E volved M agmatism D.2.1 Fractional C rystallization F ractional crystallization is a commonly invoked pro cess t hat causes evol ution in the composition of a magmatic system. Indeed, fractional crystallization can help explain many of the patterns seen in the geochemistry of the Michipicoten Island Formation . T his effect can most clearly be seen in the change in the Eu anomalies (fig. con) . Eu is unlike the rest of the REE because it can be both trivalent (the norm for REE , a 3 + charge ) and divalent (a 2 + charge) . This divalent nature allows E u to substitute for Ca in plagioclase. This allows us to understand how much plagioclase crystalliz ation has occurred between melts . T he least evolved (most mafic , Mg+Fe rich ) unit, the South Shore Member, has a small Eu anomaly in the chondrite normalized diagram , while the most evolved (most silicic or most Mg+Fe depleted ) unit, the Davieaux Island Me mber, has the strongest negative Eu anomaly (fig. 12 ) . The accumulation of plagioclase can explain why the Cuesta flows have higher Eu concentrations th a n samples from the Cha nnel Lake Member that have similar weight percent SiO 2 . This is supported petrographically where we observe that there is plagioclase accumulation within the C uesta Members. The Cuesta M ember also appears to have an accumulation of Fe - Ti oxides, which expl ains the elevated Cr concertation (fig. 9 ). Fractional crystallization of monazite or allanite can also possibly explain why the Davieaux Island Member is de p leted in the LREE compared the diorites of the Channel Lake Member (Miller and Mittlefehldt, 1982) . T he trend l ine s observed are complex and thus require modeling to further explain the variations. Due to the evolved nature of the Michipicoten Island Formation , Rhyolite MELTS was chosen as the preferred modeling program (Gualda et al., 2012) . Models where run at a wide range of internal compositions (TOR0000S1, TOR0000S2, TOR0000RC, and TOR0000RJ) with variable wate r concentration (0.5% to 5%) and diff er ent o xygen fugacities ( QFM 1.5 to - 0.5 - 1). Models were also run under variable pressure conditions (0.2 to 10 kbar). MELTS models were unable to fully reproduce the entire trend line for all the major element 39 compositional spaces (fig. 18 ). The models do get clos er to reproduction of the liquid line of the d escent under lower pressures (<2.5 kbar), low water concentrations (<2 wt. % ) and oxidizing NNO = 0). The failure of the MELTS models to accurately reproduce the entire trend lines seen in the majo r elements would argue that pure fractional crystallization is not the sole process contributing to the evolution of the Michipicoten Island Formation . Thus, we will examine other methods for magma evolution i.e. crustal anatexis and liquid Immiscibility. D.2.2 Crustal Anatexis Another commonly invoked method by which magmas can evolve is th r ough crustal anatexis. With in the MCR , this process has been most commonly used to explain the highly evolved nature of rhyolites , and their corresponding extremely n on - radiogenic i values (Nichols on, 1990; Vervoort and Green, 1997; Vervoort et al., 2007) . These extremely non - i values are interpreted as coming from the old Archean crust. If this were the case, then we should see a correlation d i i and increasing SiO 2 ; this is not observed within the Michipicoten Island Formation . Indeed, the Davieaux Island r hyolite ( the most silicic sample ), i and i . In contrast s ample TOR0000 S 1 from the South Shore member is the most mafic but has non - radiogenic i and i values. The lack of correlation does not rule out c rustal anatexis as a w hole . It simply suggests that if there was a significant amount of anatexis , then it must come from a younger , more radiogenic source. One such possible radiogenic source are the previous flows from the e arly and m ain stages of MCR volcanism. By the end of the m ain stage of volcanism , the volcanic pile has reached an approximate thickness of 20 km between the Mich ipicoten Island Fault and the Keweenaw fault. This provides a potential source of anatect ic melt . From the numerous melting experiments that have been conducted , the amount of water in the system has the most control over the final composition of the melt (Helz, 1976; Spulber and Rutherford, 1983; Beard and 40 Lofgren, 1989) . Th y et al. (1990) compiled data from many of these experiments to help explain the origin of some of the rhyolites found in Iceland. Using these insights we can observe in the Al 2 O 3 vs. SiO 2 plot t hat the h igh S i O 2 sample from the Michipicoten Island Formation overlap with partial melts of a basaltic source ( f ig . 19 ) (Thy et al., 1990) . Melting of previous MCR volcanic material has been u sed to explain the origin of isotopically primitive rhyolites in the Portage Lake Volcanics (Nicholson, 1990) . Thus, it seems l ikely that anatectic melts derived from a partial melting of a basaltic source contributed to the Michipicoten Island Formation geoche mical evolution. D.2.3 Liquid Immiscibility Magma unmixing th r ough immiscibility results in the separation of an iron r ich ferrobasalt and a rhyolite. Experimental results show that a point of liquid immiscibility can occur once the magma reaches 90 - 95% crystallization and temperatures of 1,010 - 1,040 C ° (Dixon and Rutherford, 1979; Philpotts, 1982; McBi rney, 1996) . The partition ing of elements between the two magmas leaves a pronounced geochemical signature. The ferrobasalt will be enriched in the REE and HFSE , due to its more depolymerized nature, l eaving the rhyolite depleted in these elements (Hess, 1971) . The depletion of REE and HFSE in the rhyolitic melt is the opposite of the n ormal fractio nal crystallization process , which typically concentrates REE and HFSE in the rhyolitic melt . Based on the partitioning of REE and HFSE , we would expect the Davieaux Island Member rhyolite to have low REE and HFSE concentration s if magma unmixing is playin g a role. Since the Davieaux Island Member is the most enriched in terms of the HFSE and HREE Michipicoten Island Form ation (fig. MPx, 12 ) . 41 D.3 Temporal E volution of the Michipicoten Island Formation The geometry and structure of Michipicoten Island allowed for the preservation of a nearly complete stratigraphic sequence of units. This has exposed a petrologically and g eochemically divers e record of a sili cic volcanic center, and sampling of this stratigraphy allows for the examination of its evolutio n. In this section we will describe the magmatic history of the Michipicoten island formation based on the different membe r s stratigraphic position. The Michi picoten Island Formation volcanic system began with the Cuesta Lower and Upper F lows. These first two pulses of magmatism carried a relatively large abundance of phenocryst s (CPX+PLG+OX ± BIO) (fig. 4 ). We interpret thi s as resulting from an incipient and inefficient magmatic plumbing system. In these early flows , magma stalled within the crust for long enough to generate abundant phenocryst s and glomerocryst s . As the system evolve d into the Channel La ke Member , phenocry st phases are less abundant (fig. 4 ) . The loss of phenocryst s is accompanied by a general trend from more evolved (~67 wt. % SiO 2 ) towards less evolved - up section (~ 55 wt. % SiO 2 ). This would seem to indicate that melt is spending less time within the cru st before eruption . Thus, either the magmatic plum b ing system is becoming more efficient , or the magmatic flux is increasing. The Quebec Harbor Member represents a break in the general progression of decreasing SiO 2 with t ime, jumping from ~55 wt. % SiO 2 at the end of the Channel Lake Member to ~67% wt. SiO 2 in the Quebec harbor member (fig. 7 ) . The change in geochemistry is also marked with a change in the petrology, with generally aphanitic flows seen throughout the Chann el Lake Member , shifting to plag i oclase - ph y ric flows in the Quebec Harbor Member (fig. 4 ) . Examining a Cenozoic flood basalt sequence in East Africa, Krans et al. (2018) observe d that plag ioclase - ph y ric flows often occur after a hiatus in volcanism , and this w as interpreted to be the result of plagioclase mobilization from the magma mush after recharge. While some anhedral plagioclase crystal s may have been pick ed up this way, the lack of a sharp increase in the Eu concentration between the high SiO 2 Channel Lake samples and the Quebec Ha r bor 42 Member suggest that accumulation , rather the majority of the plagioclase crystalized from the liquid (fig. 10 ) . The Quebec Harbor member represent s a period of reduce d f lux , allowing for the crystallization of plagioclase crystals pre - eruption. After the low flux period , which produced the Quebec Harbor member , the South S hore member is produced. This unit has returned to the aphanitic texture seen in the Channel L ake M ember (fig. 4 ) . The geochemistry also appears to have returned to the Channel Lake - like system as well . This suggest s that the magmatic flux has incr eased from the Quebec Harbor Member. The seemingly rapid change from the evolved geochemistry of the Q uebec Harbor Member back to a system that is more similar to the Channel Lake Member , suggest s that the more efficient magmatic plum b ing system created by the e nd of the Channel Lake Member remained intact despite the slowdown in flux between the Chan n el L ake Member and the South Shore Member. Unlike the Channel Lake Member, it appears that the m agmatic system has become stabilized in TiO 2 (i.e. unchanging with SiO 2 ) (fig. 6 ) . This is the first evi de nce of the Michipicoten Island Formation becoming buffered and i ndicates the Ti must be highly compatible in the solid phase (Lee et al., 2014) . With the eruption of the last of the South Shore Member the volcanic system shuts down . The formation of the Davieaux Island Member represents the last gasp of this magmatic system. T he exact relationship between the Davieaux Islan d Member and the South Shore Member is unclear due to the lack of a contact between the units . However, we do know that less than 1 million years p assed between eruptions , because the West Sand Bay Member is dated at 1084.4 Ma , ( below the Quebec Harbor Mem ber ) and the Davieaux Island Member 1083.5 Ma. The shift fr o m basaltic andesites in the South Shore Member to r hyolite in the Davieaux Island Member clearly represents a shift in the nature of the magmatic system. Previously when the Michipicoten Island Fo rmation contained high silica products (e.g. Quebec Harbor M ember ) it was marked with an increase in the abundance of phenocryst s , but the 43 Davieaux Island rhyolite is conspicuous ly devoid of phenocryst s (fig. 4 ). A mechanism for generating crystal poor rhy olites is compression and melt extraction (Bachmann and Bergantz, 2004; Hildreth, 2004) . This would seem to be further supported by the geochemistry in which the Davieaux Island Membe r seemingly falls off the Zr trend line formed by the other Michipicoten Island Formation samples (fig. 10 ) . This signature can be generated by melt initially following the typical liquid line of descent with increasing Zr with increasing SiO 2 , t hen once the melt reaches z ircon saturation and 50% crystallinity it can be compress ed and extraction will cause the erupted of a rhyolite with depleted in Zr (fig. 20 ) (Deering and Bachmann, 2010) . D.4 Temporal C ontext of the Michipicoten Island Formation in the F ramework of MCR M agmatic S tratigraphy M agmatism within the MCR has been broken into four stages: Initiation, early, main, and late , each with some unique geochemical characteristic . The initiation stage is composed of ultramafic to mafic intrusive s related to melting from the impacting plume and thus drawing any comparison to the late stage magmas would be futile (Heaman et al., 2007; Miller and Nicholson, 2013) . The early s tage is largely composed of basaltic material with some studied evolved melts found in the North Shore Volcanic Group (NSVG) (Dosso, 1984; Vervoort and Green, 1997; Vervoort et al., 2007) . Magmatism from the hiatus stage is rar e and is only found at Mamainse Point. The Mamainse Point formation is also the m ost well c haracterized geostratigraphic section within the MCR volcanics , making it a useful comparator to the Michipicoten Island sample suite. Basaltic volcanism from the main stage can be found throughout the rift but silicic volcanism is more prevalent on the western ar m of the MCR. This study is the first rigorous geochemical study of late stage volcanism within the MCR . I t is thus important t o understand the similarities in geochemistry between the late stage magmas and 44 previous magmatic stages becaus e they can provide insight into how the Michipicoten Island Formation formed . D.4.1 Early Stage Evolved magma s are rare during the early stage of magmatism in the MCR. The evolved melts from the early flows of the NSVG ha ve a broadly similar trace element patter n to the Michipicoten Island Formation , especially in the highly incompatible elements , although the slop e across REE is steeper than the Michipicoten Island Formation samples (fig. 21 ) (Vervoort and Green, 1997) . It should be noted that a steep slope to the REE is a common feature of the early stage basalts (Shirey et al., 1994; Nicholson et al., 1997) . Isotopically , the melts prod uced from this stage have only a slig htly non - radiogenic values of ca. - (Ver voort and Green, 1997; Vervoort et al., 2007) . Evolved melts during this time migrate d though the cold continental crust with little interaction , and evolve by fraction crystallization and assimilation of previous MCR mate rial or high Sm/Nd crust (Vervoort and Green, 1997) . D.4. 2 Hiatus Stage There is very little volcanism preserved from the hiatus stage . Dating of silicic clast s from the Copper Harbor Conglomerate in the Keweenaw P eninsula by Davis and Paces (1990) indicate that during the hiatus stage the re was some silicic volcanism in the western arm of the MCR (fig. 2 ) . The only place where hiatus volcanism is preserved in their origi nal flows is within the Mamainse Point Sequence. The flows from the Mamainse Point Sequence have been broken into 8 distinct groups based on their geochemistry (Klewin and Berg, 1990) . The flows from the hiatus stage are referred to as Group 5 . The Group 5 lavas from Mamainse Point can be broken into three subgroup s 5 a, b and c (Klewin and Berg, 1991). These basalts appear to have different trace element patterns, with Group 5 b having the lowest enrichmen t in trace elements (Fig. 21 ). Despite this, Klewin and Berg (1991) were able to show that the Group 5 a magmas 45 co uld be reproduced by simple mixing of ~80% 5b melt and ~20% rhyolitic melt, which the authors use as their crustal melt analog. The Group 5 c ma gmas are more compl ex , requiring both fractional crystallization and crustal contamination to produce the trace e lement enrichment. The Group 5 c trace elements f o llow a very similar trace element pattern to the Michipicoten Island Formation (fig. 21 ). Isot opically the Group 5 basalts from Mamainse Point appear to be quite different from those of the Michipicoten Isla nd Formation with more non - - isotopic data from the Group 5 lavas at Mamainse P oint led Shirey et al. (1994) to interpret the source for these lavas was the primiti ve mantle (i.e. p lume ) with variable amount of continental crust. D.4. 3 Main Stage L y ing beneath the late stage volcanic of the Michipicoten Island Formation is the Quebec Mine Member basalts, which are texturally similar to the m ain stage lavas found at Mamainse Point (Annells, 1974) . The m ain s tage lavas at Mamainse Point have been bro ken into 3 groups : G roups 6, 7 and 8. Group 8 type lavas are only found at Mamainse Point and ha ve a geochemical signature unlike any other volcanics found in the MCR (Shirey et al., 1994; Nicholson et al., 199 7) . Trace element patterns are sim i lar between the Group 6 and 7 but Group 6 has higher trace elemen t concentrations ( 21 ) . Group 6 trace elements patterns , in particular , resemble the trace element pattern of the Michipicoten Island Formation in the more compatible trace elements (fig. 21 ). Isotopically , Group 6 and 7 lavas lie between the primitive mantle and depleted mantle components , with Group 6 closer to the primitive mantle and Group 7 - (fig. 13 ) . Shirey (1994) interpreted these isotopic data to indicate that the m ain s tage volcanism is the result of variable amounts of the plu me and depleted mantle ; this has been supported by other authors working on other main stage volcanic sequences (Paces, 1988; Paces and Bell, 1989b; Nicholson et al., 1997) . The 46 majority of the Michipicoten Island samples plot near the Group 6 and 7 basalts from Mamain se Point, though are farther off the mantle array (fig. 13 ) . E volved magmatism from this stage is has been predominantly studied in two formations , t he NSVG on the western shore of Lake Superior and the Portage Lake Volcanics (PLV) in the Keweenaw Peninsu la. The trace element patterns of the main stage NSVG icelandite s appear similar to the samples from Michipicoten Island , although the rhyolite taken on in Zr, Hf, Sm, Nd and LREE that the Davieaux Isla nd Member exhibits (fig. 21 and 11 ). In the NSVG the rhyolitic and granophyric magmas have highly non - i values ( - 15 to - 2) (Vervoort and Green, 1997; Vervoort et al., 2007) . The intermediate magmas ( iceland ites and andesites) from the main stage have less no n - i values ( - 9 to 0, most < - 4) (Dosso, 1984; Vervoort and Green, 1997) . The difference in isotopic signature between the rhyolites and the intermediate magmas has been interpreted to be the result of the melting of the highly non - radiogenic Archean crust producing rhyolitic and granophyric magmas, while the intermediate magma ( like the evolved melts of the main stage ) are the result of fractional crystallization with little interaction from the crust (Vervoort and Green, 1997) . Nicholson (1990) noted that many of the rhyolites in the Portage Lake Volcanics, found in the Keweenaw Peninsula, fall broadly into two categories: type 1 and type 2. These magma types were defined on the basis of petrologic differences. Most notably, type 2 shows the presence of quartz phenocrysts, whereas in t ype 1 magmas quartz is only present in the groundmass. From a petrographic standpoint, the Davieaux Island Rhyolite would appear most similar to the type 1 rhyolites. From a t race element perspective the Davieaux Island Member is most similar to the type 1 , with a more sloped REE pattern and not as pronounced Eu anomaly in comparison to the t yp e 2 rhyolites (Nicholson, 1990) . The Davieaux Island Member deviates i values : - 13.3 to - 15.9 , while ty i values between - 0.3 to - 47 of 2. 4. The type 1 Rhyolites have been interpreted to have been the result of fra ctional crystallization of the Portage Lake Volcanics and melting of the previous basalt. The late stage volcanics Michipicoten Island Formation , from a trace element perspective , appear most similar to the Group 5 basalts from Mamainse Point (erupted during the hiatus stage ) and icelandite s from the NSVG during the main phase of magmatism . Isotopically , however , the Michipicoten Island Formation samples are different from the Group 5 basalts and are most similar to main stage basalts from Mamainse Point ( G roups 6 and 7). A common interpretation for evolved magmatism with highl y incompatible trace enrichments and i values is that it is the result of melting and assimila tion from previous MCR volcanics (Nicholson and Shirey, 1990; Vervo or t and Green, 1997; Vervoort et al., 2007) . D. 5 Hybridization of L ate S tage M agmatism in the MCR The formation of intermediate lavas, like those observed on Michipicoten Island, may be the result of mixing of a rhyolitic melt with a basaltic melt . As discussed above , the Michipicoten Island Formation trace element patter n s look similar to those of the Gr oup 5 magmas from Mamainse Point . Klewin and Berg (1991) showed that this trace element pattern can be repr oduced through fractional crystallization and rhyolite mixing . This has been physically observed at Mamainse Point where basalts appear to be intermingling with rhyolites (Matthews and Rooney, 2009) . A re quirement for any silic ic melt that has mix ed with the Michipicoten Island Formation is that the rhyolite must contain radiogenic Nd and Hf isotopic values as well as non - radiogenic Sr isotopic value s (unlike the Archean continental crust) ; otherwise mixing of the rhyolite would cause a crustal contaminat ion signature . If the rhyolites were derive d from a source that was from previous MCR material, then mixing with the rhyolite might cause a perturbation towards the PM source , since most of the MCR flood basalts have a strong PM isotopic signature. Isotopi cally , it is impossible to tell the difference between direct crustal anataxis of earlier MCR material and m ixing of rhyolites that are formed from the same source. 48 Petrographically , however , there is evidence for magma mixing . T he spherical mesostasis obs erved in the Chan n el Lake Member discussed in Bacon ( 1986 ), which are interpreted to be undercooled mafic melts that are mixing/mingling with the host evolved magmas . Thus, it i s apparent th at some amount of magma hybridization is playing a role in the formation of the late stage magmas on Michipicoten Island. D. 6 Mantle S ource s of L ate M agm atism in the MCR Traditionally there has been discussion of four distinct isotopic reservoir s contributing to the MCR volcanics : the Primitive Mantle (PM) , the Depleted Mantle (DM M ) , Sub Continental Lithospheric Mant le (SCLM) , and the Archean Continental Crust (CC) (Nicholson and Shirey, 1990; Shirey et al., 1994; Shirey, 1997; Vervoort and Green, 1997; Ver voort et al., 2007) . The Primitive Mantl e isotopic composition for the MCR within the Nd system is thought t o be chondritic, which has been interpreted as melting from an upwelling thermochemical anomaly such as a plume source (N icholson and Shirey, 1990; Shirey et al., 1994; Shirey, 1997; Vervoort and Green, 1997; Wirth et al., 1997; Vervoort et al., 2007) . The chondrit ic nature of the plume has also been applied to the Sr isotopic system (Nicholson and Shirey, 1990; Shirey et al., 1994) . Since t he Nd and Sr isotopic systems both indicate the plume source is chondritic then it would stand to reason that the Hf isotopic composition of the plu me would be chondritic as well. In this study we use the chondritic Nd and Hf isotopic values from Bouvier e t al. (2008) an d Sr isotopic values from Workman and Hart (2005) . From current work being conducted on the Mamainse Point Seq uence (not part of this project) , the Pb isotopic data radiate out of a single point . This most likely represents the plume component , since th is co mponent is the only one that all the Mamainse Point basalts share (Rooney et al., 2018) . The depleted MORB mantle is another source that is thought to be contributin g as a source of magmatism in the MCR , although this source seems to be primarily restricted to the m ain s ta ge of volcanism (Shirey et 49 al., 1994; Nicholson et al., 1997; Miller and Nicholson, 2013) . This source is defined by its radiogenic Nd and Hf , and unradiogenic Sr and Pb isotopic chara cteristics. For this reservoir we use the d epleted mantle Nd isotopic value from Bennett (2003) , Hf fro m Blichert - Toft et al (1997) , Sr from Workman and Hart (2005) and Pb from Zindler and Hart (1986) . The continental crust is a more variable so urce reservoir . Underlying the MCR in the Michipicoten I sland region is the Wawa Greenstone Belt . As discuss ed in the in the crustal anatexis section , the contamination derived from old Archean crust would not have to be very significant to alter the isoto pic signature of the Michipicoten Island Formation , and thus must be considered. The Wawa Greenstone Belt is a sub - province of the Superior C raton and is thought to have formed as a result of s ubduction accretion complexes (Polat and Kerrich, 2000) . The Wawa Greenstone Belt is the closest pre - rift crustal formation to Michipicoten Island. Despite an initial positive Nd - f isotopic signature during formation, isotopic ingrowth in the Wawa Greenstone Belt has resulted in s trongly negative Nd and f values due to low Sm/Nd and Lu/Hf ratio s (fig. 13 ) (Polat and Münker, 2004) . The Pb isotopic system is much more difficult to constrain since there has been no whole rock Pb isotopic st udy of the Wawa Greenstone Belt . In an effort to try to co nstrain our crustal end member in terms of Pb , we examine the Mamains e P oint sequence (Shirey et al., 1994) . The array created by the Group 5 magmas, which are considered to be the most cr ustally contaminated of the Mamainse Point basalts , demonstrate the influence of the continental crust . W e can place our crustal endmember at the end of this array opposite of the plume source (Rooney et al., 2018) . The role of the lithospheric mantle in the generation of the Michipicoten Isl and Formation is more difficult to assess. The complete lack of mantle xenoliths from the region means that there are no direct measurements of the SCLM in the region. C arbonatite magma from the Seabrook L ake Carbon at ite , which has been interpreted as prim ary melts from the SCLM, shows Nd i of +4.5 and an initial 86 Sr/ 87 Sr of 0 .70265 (Bell and Blenkinsop, 1987) . This differs from the isotopic characteristic proposed by Shirey et al. (1994) . Recent work on the trace 50 elements of m etabasalts of the Co ldwell Complex indicate s that they are the result of melting of a metasomatized SCLM . Two samples taken from the outer metabasaltic unit have negative Nd values (Good and Lightfo ot, 2019) . Our project place s isotopic constraints (Nd, Hf, Sr, and Pb) on the SCLM by utiliz ing samples from the Wolfcamp Basalts at the interior of the Coldwell Complex , which have positive Nd i values (Rooney et al., 2018) . W e will use those preliminary results to constrain our SCLM end member . With these main components defined we can now attempt to a ssess their contribution to the Michipicoten Island Formation . To a ssess what the relative contribution of each source (PM, DMM, SCLM , and CC) the most direct approach would be to create a four co mponent unmixing model to quantitatively assign the contribu tion of each component . The r eason we were unable to conduct such a model is due to the source s of magmatism not being as well defined for the Midcontent R ift system as they are for other rift sys tems . Thus , c reating a numerical unmixing model would be pro ne to significant error because the solution would lack an illustrative topology of the mixing relationships. For our purpose s , it is important to examine such mixing relationships and the relativ e position of the Michipicoten Island samples in relation to the endmembers noted above . To address this, we create ternary mixing models w hereby t w o reservoirs where chosen ( A and B ) and a mixing line was solved for iteratively between the two (eq n . 5.2 and 5.3 ) . Then, for every iteration of that mix , a new mixing line was created between the A - B mix fraction and a new component C. From the isochron diagrams we know that the Nd and Hf isotopic systems are the most reliable in terms of closure (fig. 17 ) . The source for the Nd and Hf isotopic systems also has the mo st well - defined magmatic endmembers, which make the ( 143 Nd/ 144 Nd ) i vs. ( 176 Hf/ 177 Hf ) i plot ideal for the mixing models (fig. 22 ) . From mixing these source s in Nd, Hf isotopic space we can ob serve that most samples lie within the DMM, CC (continental crust) , and SCLM mixing field . The two South Shore Members lie just on the outside of the mixing field with TOR0000S1 near the CC - SCLM line , and TOR0000RQ near the DMM - SCLM line. The DMM might hav e had a 51 less radiogenic signature 1.1 billion years ago th a n our e quations predict which , would explain why TOR0000RQ is not captured in the mixing field . As for TOR0000S1 , there are more non - radiogeni c examples of the Wawa Greenstone Belt that could be contributing as the crustal source. The units that lie within the mi xing fields that lie above the mantle array can be attributed to < 2% mixing of the continental crust and samples lying below the mantl e can be attributed to <2% mixing of the subcontinental lithospheric mantle respectively . A n issue with the 143 Nd/ 144 Nd v s. 176 Hf/ 177 Hf isotopic diagram is that the PM source can be replicated through the mixture of DMM, SCLM , and CC , thus other isotope sp ace s must be analyzed to assess whether the plume source is present and whether the samples plot within the same mixing f ield as in the ( 143 Nd/ 144 Nd ) i vs. ( 176 Hf/ 177 Hf ) i plot . Additional ternary mixing models were created in ( 206 Pb/ 204 Pb) i vs. ( 143 Nd/ 144 Nd) i and ( 206 Pb/ 204 Pb) i vs. ( 176 Hf/ 176 Hf) i plots , which unlike in the ( 143 Nd/ 144 Nd) i vs. ( 176 Hf/ 176 Hf) i plot , the PM comp CC mixing field due to the more radiogenic 206 Pb/ 204 Pb signature of the PM. From these diagrams , we observe that the Michipicoten Island samples plot near the PM source and outside of the DMM, SCLM and CC field. These two facts mean that the primitive mantle is a required magmatic source for the Michipicoten Island Formation. Our model may not be able to give us exact numbers on how much of each source contributed to the Michipicoten Island Formation samples, but they do illustrate how much control the different components have. Because the concentration of elements in the contine ntal crust and the lithospheric m antle are so much higher than the PM and the DM M reservoirs, the fact that most of the samples lie near the DM M and PM components indicates that they must be contributing the largest proportion of melt. T he depleted mantle signature in the late stage magmas of the Michipicoten Island Formation is an interesting outcome since this source indicates decompress ion of the asthenosphere must have continued during this late stage of magmatic activity. Such an interpretation has imp lications for geodynamic models of rift development and the eventual failure of the MCR. 52 D. 7 Implication for R ift F ailure The origin al interpretation of the Great Lakes International Multidisciplinary Program on Crustal Evolution (GLIMPCE) seismic lines resulted in the interpretation that MCR failure was due to far - field effects of the Grenville orogeny (Cannon and Hinze, 1992; Cannon, 1994) . Reinterpretation of the GLIMPCE lines by Stein et al. (2015) using numerical stepwise structural restoration models call ed into question whether the MCR failed due to orogenesis , instead suggesting that successful rifting of Amazonia and Laurentia resulted in the remov al the stress required for continued continental rifting of the MCR after the early stage of volcanism. Swanson - Hysell et al. (2019) point out that a n Amazonia - Laurentia rifting model is difficult to reconcile with the timing of the Ottawa phase of the Grenville Orogeny and that the Amazonia Laurentia rift would have had to have formed and then shortly afterwards b een inverted . Instead , Swanson - Hysell et al. (2019) us ed dating of the unconformities , which are refer red to - an indicator of when the transition from active rifting to thermal subsidence began . Swanson - Hysell et al. (2019) conclude d that the cessation of act ive rifting at 1091 Ma is correlative with the beginning of the Ottawa phase of Grenville orogeny . Dating of volcanic bodies from the Adirondack highlands indicates that the Ottawa phase of the Grenville Orogeny had begun by 1080 ± 4 Ma (Chiarenzelli an d M cLelland, 1991; McLelland et al., 2001). T he DM M signature within the Michipicoten Island Formation suggest s that thinning of the lithospheric mantle continued until 1083 Ma - this is within error of beginning of orogenesis. This would seem to indicate tha t these two events are likely correlated and lends additional evidence to the Cannon (1994) model for the MCR failure as a result of far - field effects of the Grenville orogeny. 53 Conclusion The Michipicoten I sland F ormation is the youngest exposed volcan ic sequence from the final stage of volcanism within the faile d Midcontent rift . A s such it preserves the conditions during final stages of activity within this rift. Lavas within the Michipicoten I sland F ormation are more evolved than would be anticipat ed from a terminal oceanic rifting env ironment . W ithin an advanc ed oceanic rifting environment , such as Afar (East Africa), the silicic magmatism is accompanied by primitive basaltic magmatism. Within the Michipicoten I sland F ormation , we lack primitive basaltic magmatism and only have examples of evolved magmatic acti vity. Our major and trace element chemistry f or m an array from relatively undifferentiated to differentiated compositions extending from basaltic andesites to rhyolites . The continuum of compositions is unusua l for rifting environments , and likely reflects mixing of rhyolitic and basaltic ma g mas. The major and trace element data of the Michipicoten Island Formation resembles other units within the MCR - notably hybridized Group 5 basalts from Mamainse Point . Ho wever, our new isotopic data show significant d iffe re nce s between the Michipicoten I sland F ormation and other evolved magmatism within the MCR . The se older evolve d magmas appear to be the result of mixing between the p rimitive mantle and old continental crust ; the Michipicoten I sland F ormation isotopic ally appear s to resemble the main stage of volcanism that was the result of mixing between the depleted mantle and the primitive m antle during a period of extension . Our results show that these sources continued to contribute to the final stages of magmati sm within the MCR . Importantly , Archean continental crust was no longer significa nt ly contributing to the final stage , suggesting that Archean crustal material may not be present at this late stage of rifting. . . This implies that magma chambers of the Mi chipicoten I sland F ormation were located within the previous volcanic units , w h ich ma y have contributed chemically but would be difficult to resolve isotopically because of their similar composition. Previous episodes of evolved magmatic activity within th e MCR tended to coincide with hiatuses in magmatism. During the se hiatus stage s, extension is thought to have waned and lavas from this period appear to be dominated 54 by melts derived from a plume source mixed with the continental crust ; little contribution from the depleted upper mantle is observed . Unlike the se older hiatu s events, th e Michipicoten I sland Formation appears to continue to have the same isotopic characteristics as the previous m ain stage of volcanism . The implication of this observation is t hat decomp ression melting of the depleted upper mantle continued during this final stage of magmatic activity. Such an observation has a profound impact on geodynamic models for the development and failure of the MCR , as it implies plate thinning continued to ca. 1083 Ma. Future work will examine the disconnect between evidence within the crust of a much earlier cessation in the manifestation of extension, and the results presented here that require continued thinning of the lithospheric mantle. 55 APPENDI CE S 56 APPENDIX A FIGURES 57 Figure 1: Gravity Anomaly Map Gravity anomaly map modified from Stein et al. (2015) . Arrow indicating the Michipicoten Island gravity high. 58 Figure 2: MCR Stratigraphy Stratigraphic columns of magmatic and sedimentary units found within the MCR. Dates in bold are from Fa irchild et al. (2017) and other dates are from Palmer and Davis (1987) . Upper right map is a magnetic anomaly map of the Lake Superior region . Figure modified from Fairchild et al. (2017 ) . 59 Figure 3: Geologic Map of Michipicoten Island Geologic map of Michipicoten Island . Figure modified from Fairchild et al. (2017) origi nal map from Annells (1974) . Pink pentagons are the s ample locations. 60 Figure 4 : Photomicrograph s Crossed polarized light images taken at 2X magnification of samples from the Michipicoten Island Formation. a . Davieaux Island Member. b . South Shor e Member. c . Quebec Harbor Member. d . Channel Lake Member. A m agmatic inclusion is in the center of the imag e. e . Cuesta Upper Flow. The image highlight s a glomerocryst f . Cuesta Lower Flow. 61 Figure 5: Classification Diagrams a. Total Alkali Silica classification diagram (Le Bas et al., 1986) and alkaline v. sub - alkaline discrimination line (Irvine and Baragar, 1971) . b . Discrimination diagram of tholeiitic v. calc - alkaline (Irvine and Baragar, 1971) . FeO* is the total iron calculated as FeO. 62 63 Figure 6: Major E lements D iagrams Major el ements diagrams of the Michipicoten Island Formation. Key is ordere d by stratigraphic position 64 65 Figure 7 : Flow E volution D iagrams Flow evolution diagram s showing the change in the major elements with relative stratigraphic position. 66 Figure 8: Large Ion Lithophile Element Diagrams Large ion li thophile element v. silica 67 Figure 9: First Row Transition Elements Diagrams First row transition elements v. silica 68 Figure 10: High Field Strength Element Diagrams High field strength elements v. silica 69 70 Figure 11: Primitive Mantle Normalized Trace Element Diagram s Primitive mantle normalized trace element. Faded purple region indicating the range of the Michipicoten Island Formation. Samples ordered by stratigraphic position and normalized to McDonough and Sun (1995) 71 72 Figure 12: Chondrite Normalized R are Earth Element Diagram s Chondrite normalized rare earth element diagram s . Dashed line in the Quebec Harbor Member graph indicate sample was taken from the altered section as mapped by (Annells, 1974) . Faded purple region indicating the range of the Michipicoten Island Formation. Samples ordered by stratigrap hic position and normalized to McDonough and Sun (1995) . 73 Figure 1 74 i Digram i v. i plot showing the Michipicoten Island Formation samples with Groups 5, 6 and 7 from Mamainse Point. Also shown are carbonatite magmas and alkaline basalts from the Coldwell Complex. Wawa Greenstone Belt (WSB) data is from Polat and Münker (2004) and is added to the plot because it is the closest Archean crustal material to Michipicoten Islan d. All data is age corrected to 1 100 Ma. Mantle array from Chauvel et al. (2008) . Deplete d MORB mantle (DMM) composition from Bennett (2003) and Blichert - Toft et al. (1997) . Primitive Mantl e composition is chose n to be chondrit ic and thus must lie at Red circles indicate samples where their original isotopic signature were likely perturbed by alteration . 75 Figure 14: ( 87 Sr/ 86 Sr) i i Diagram ( 87 Sr/ 86 Sr ) i v . i plot showing the Michipicoten Island Formation samples with carbonatite magmas and alkaline basalts from the Coldwell Complex. As well as Portage Lake Volcanic (PLV) samples from Nicholson and Shirey (1990) . Archean crustal field created from Sr isotopic data from Turek et al. (1982) and Nd isotopic data from Polat and Münker (2004) . Red circles indicate samples that their original isotopic signature was likely perturbed by alteration . 76 Figure 15: Pb Isotope Diagram s Pb isotope plots showing the Michipicoten Island Formation samples with Groups 5, 6 and 7 fr o m Mamainse Point. Also shown are alkaline magmas from the Coldwell Complex. Location of 77 Figure 1 5 the primitive mantle is based on the conver gence the Mamainse Point data. Red circles indicate samples where their original isotopic signature were likely perturbed by alteration. a. 207 Pb/ 2 04 Pb versus 206 Pb/ 20 4 Pb. b. 208 Pb/ 204 Pb versus 206 Pb/ 204 Pb. Location of the continental crust (CC) Pb isotopic ratio based on Mamainse Point Group 5 , which is interpreted to be a mix between PM and CC sources. DMM isotopic composition based on Zindler and Hart (1986) . 78 Figure 16: Isocon D iagrams Isocon diagrams. These diagrams show the element mobility by plotting the element concentration unaltered sample on the x axis against an altered sample but one of similar origin on the y axis . Element concentrat ions were multiplied by a scaling factor to bring them to within the plot region because element concentration s were vastly different between members. The b lack line is the slope and is equal to 1 , thus if th e concentration in the alter ed sample is the sam e as the unaltered then it will fall on this line. Blue lines indicating the ± 10% of the slope equal to 1. 79 Figure 17: Isochron Diagram s Isochron p lots of the Michipicoten Island Formation. Blue line is t he 1100 Ma chondritic isoch ron. Y axis values ar e the measured isotopic values , X values are calculated (e.q. 1.2, 2.2, 3.2, 4.4). a . 87 Sr/ 86 Sr versus 87 Sr/ 86 Rb b. 176 Hf/ 177 Hf versus 176 Lu/ 177 Hf c. 143 Nd/ 144 Nd versus 147 Sm/ 144 Nd d. 208 Pb/ 204 Pb versus 232 Th/ 204 Pb e. 207 Pb/ 204 Pb versu s 235 U/ 204 Pb f. 207 Pb / 204 Pb versus 235 U/ 204 Pb 80 Figure 18: MELTS M odels Liquid lines of descent for a range of MELT S model s . Conditions for th es e m odel s w ere p ressure set at 2 kbar , oxygen fugacity set at NNO = 0 and the starting composition was sample T OR0000S2. Model vari ed by concentration of water added. 81 Figure 19: Chemical Comparison of Michipicoten Island Formation Samples to Experimentally Produced Liquids Al 2 O 3 versus SiO 2 plot s howing the Michipicoten Island Formation sample s compared to exper imentally produced l iquids from melting of basalt under different h ydration conditions . . A. Lowest H 2 O field. B. Moderate H 2 O field C. Highest H 2 O field. Figure modified f rom Thy et al. (1990) . 82 Figure 20: Liquid E xtraction Diagram Figure from Deering and Bachmann (2010) showing the behavior of Zr during liquid extraction. 83 Figure 21: Primitive M antle N ormalized Comparison Diagrams Primitive mantle normalized trace element diagra m s from different re gions of the MCR. Faded purple region indicating the range of the Michipicoten Island Formation . Samples ordered by 84 Figure 2 1 stratigraph ic position and normalized to primitive mantle (McDonough and Sun, 1995) . a . Ma main se Point s amples from the main stage of volcanism di vided based on the work of Klewin and Be rg (1991) and new trace element concentrations from Rooney (in progress) . b . Mamains e Point samples from the hiatus of volcanism divided based on the work of Klewin and Berg ( 1991) and new trace element concentrations from Rooney (in progress). c . North Shore Volcanic Group samples from both the m a in and e arly stages of volcanism (Vervoort and Green, 1997) . Icelandite are solid lines with circles and the rhyolite is a dashed line with stars. 85 Figure 22: Ternary Magma Mixing Model Ternary m agma mixing model plots. Red circles indicate samples where their original i sotopic signature were likely perturbed by alteration. Dark blue lines are mixing lines between PM, SCLM and DMM. Light b lue lines are mixing lines between PM, SCLM and CC. Red lines are mixing lines between PM, DMM and CC . This model was created by first iteratively c al culating a primary mixing line between two endmembers, then for each i ter ation a secondary mixing line is generated between the mix and the third endmember. Each point represents an iterative calculation. Only select secondary mixing lines a re shown. a. Hf vs . Nd isotopic mixing p lot. PM, SCLM and DMM mixing field (dark blue) generated with primary mixing between DMM and PM , at every 10% . S econdary mixing of (DMM, PM) mix with the SCLM lines at every 1 0 % mix of PM, 86 Figure 2 point s a t every 1% mix of SCLM . PM, SCLM and CC mixing field (l ight blue ) with primary mixing between PM and SCLM at every 1% mix . S econdary mixing (PM, SCLM) mix with the CC lines between 99% and 90% mix of PM with point s at every 1% mix of CC . PM, DMM and CC mixi ng field (r ed ) generated by with primary mixing between DMM and PM at every 10% mix . S econdary mixing (DMM, PM) mix with the CC point s at every 1% mix . b. Pb versus Nd isotopic mixing plot. PM, SCLM and DMM mixing field (dark blue) generated by the primary mixing between DMM and SCLM point s at every 1% mix . S econdar y mixing (DMM, SCLM) mix with the P M , l ines between 0% and 3% SCLM point s at every 1% PM . PM, SCLM and CC mixing field (light blue) primary mixing between SCLM and CC point s at every 1% mix . Second ary mixing (SCLM, CC) mix with the PM, lines every 10% betwee n 100 and 10% SCLM point s at every 1% PM. PM, DMM and CC mixing field (red) generated by with primary mixing between DMM and CC point s at every 1% . S econdary mixing (DMM, CC ) mix with the PM, lines between 99 and 90% with 0% DM M point s at every 1% PM . c. Pb versus Hf isotopic mixing plot. PM, SCLM and DMM mixing field (dark blue) generated by the primary mixing between DMM and SCLM point s at every 1%. Secondary mixing (DMM, SCLM) mix with the PM , lin es between 0% and 5 % SCLM point s at every 1% PM. PM, SCLM an d CC mixing field (light blue) wi th primary mixing between SCLM and CC point s at every 1%. Secondary mixing (SCLM, CC) mix with the PM, lines every 10% between 100 % and 10% SCLM point s at every 1% PM. PM, DMM and CC mixing field (red) generated by with prim ary mixing between DMM and CC poi nt s at every 1%. Secondary mixing (DMM, CC) mix with the PM, lines between 99 and 9 5 % with 0% of DMM point s at every 1% of PM. 87 APPENDIX B TABLES 88 Table 1 : Maj or Elements C oncentrations Sample Unit Member SiO2 (%) TiO2 (%) Al2O3 (%) Fe2O3 (%) MnO (%) MgO (%) CaO (%) Na2O (%) K2O (%) P2O5 (%) LOI (%) Sum (%) TOR0000 SB Michipicoten Island Intrusive 63.58 1.08 13.11 7.13 0.15 1.62 3.85 4.43 1.81 0.24 2.79 97 TOR0 000 SD Cuesta Lower Flow 56.57 1.94 14.31 10.34 0.23 2.87 5.8 8 3.49 2.08 0.5 1.6 98.21 TOR0000 SE Cuesta Lower Flow 55.87 2.01 14.19 11 0.23 3.03 6.11 3.29 2.11 0.52 1.46 98.36 TOR0000 SF Cuesta Lower Flow 55.87 1.97 14.03 10.65 0.22 3.08 6.41 3.21 2.02 0.5 1 1.85 97.97 TOR0000 SG Cuesta Lower Flow 56.31 1.97 14.22 1 0.49 0.25 3.04 5.64 3.53 2.13 0.51 1.74 98.09 TOR0000 QQ Cuesta Upper Flow 55.76 1.61 15.47 8.22 0.13 4.16 1.3 2.08 7.17 0.52 3.22 96.42 TOR0000 QR Cuesta Upper Flow 58.23 1.54 15.23 7.94 0.11 2.8 9 2.4 2.67 5.71 0.48 2.61 97.2 TOR0000 QS Cuesta Upper Flow 59.06 1.44 15.02 7.7 0.15 1.7 5.8 3.88 1.18 0.46 3.39 96.39 TOR0000 QJ Channel Lake Member 67.43 0.53 13.94 4.11 0.1 0.5 2.9 4.78 1.73 0.1 3.6 96.12 TOR0000 QK Channel Lake Member 67.91 0.54 14.05 4.18 0.1 0.53 2.96 4.83 1.77 0.1 2.77 96.97 TOR0000 QL Chan nel Lake Member 70.73 0.52 13.65 3.84 0.07 0.68 0.33 3.75 5.08 0.1 1.06 98.75 TOR0000 QM Channel Lake Member 67.42 0.58 14.04 4.47 0.11 0.6 2.89 5.02 1.83 0.12 2.68 97.08 TOR0000 QN Channel Lake M ember 73.37 0.58 13.26 2.03 0.02 0.24 1.23 3.5 4.54 0.12 0.8 9 98.89 TOR0000 QO Channel Lake Member 67.46 0.58 14.13 4.44 0.12 0.64 2.92 4.85 1.91 0.12 2.6 97.17 TOR0000 QT Channel Lake Member 67.05 0.64 14.11 4.64 0.11 0.7 3.2 4.77 1.72 0.14 2.69 97.08 TO R0000 QU Channel Lake Member 66.81 0.61 14.06 4.62 0.11 0.68 3.09 4.68 1.67 0.13 3.32 96.46 TOR0000 QV Channel Lake Member 66.74 0.6 13.96 4.48 0.11 0.65 3.17 4.68 1.62 0.13 3.63 96.14 TOR0000 QW Channel Lake Member 66.94 0.66 14.1 4.69 0.11 0.73 3.3 4.69 1 .6 0.14 2.8 96.96 TOR0000 QX Channel Lake Member 72.75 0.51 13.52 1.61 0.03 0.17 0.71 3.31 5.38 0.1 1.68 98.09 TOR0000 QY Channel Lake Member 65.74 0.74 14.04 5.3 0.12 0.91 3.61 4.54 1.57 0.16 3.04 96.73 TOR0000 QZ Channel Lake Member 66.09 0.73 14.14 4.91 0.12 0.82 3.6 4.58 1.5 0.17 3.11 96.66 TOR0000 R1 Channel L ake Member 58.99 1.45 14.84 8.84 0.16 3.45 1.6 4.05 3.42 0.43 2.57 97.23 TOR0000 R3 Channel Lake Member 57.22 1.58 14.45 9.38 0.26 2.86 6.51 3.57 0.99 0.41 2.55 97.23 TOR0000 R4 Channel Lake Membe r 57.72 1.6 14.38 9.75 0.18 2.62 6.34 3.78 1.09 0.4 1.96 97. 86 TOR0000 R5 Channel Lake Member 62.49 1.49 13.56 8.14 0.09 1.88 3.17 2.24 4.38 0.4 1.93 97.84 TOR0000 R6 Channel Lake Member 63.14 0.98 14.3 6.28 0.14 1.21 4.47 4.14 1.33 0.25 3.49 96.24 TOR000 0 R7 Channel Lake Member 62.66 1.08 14.37 6.58 0.16 1.43 4.8 4.08 1.25 0.27 3.05 96.68 TOR0000 R8 Channel Lake Member 56.76 1.96 14.03 10.46 0.23 2.75 6.72 3.56 0.91 0.63 1.8 98.01 TOR0000 R9 Channel Lake Member 55.86 1.92 13.74 10.44 0.22 2.6 6.84 3.44 0.8 6 0.61 3.28 96.53 TOR0000 RA Channel Lake Member 56.89 1.92 13.81 10.41 0.26 2.68 6.81 3.53 0.86 0.61 2.03 97.78 TOR0000 RB Channel Lake Member 53.97 2.02 14.29 13.05 0.19 4.35 2.96 3.81 1.58 0.64 3 96.86 TOR0000 RC Channel Lake Member 56.25 1.92 13.77 10. 74 0.22 2.68 6.91 3.57 0.78 0.6 2.38 97.44 TOR0000 RD Channe l Lake Member 56.44 1.96 13.89 10.81 0.22 2.71 6.87 3.62 0.78 0.6 1.92 97.9 TOR0000 RE Channel Lake Member 56.18 1.91 13.84 10.5 0.24 2.76 6.81 3.48 0.76 0.55 2.79 97.03 TOR0000 RF Channel Lake Me mber 64.95 0.75 14.2 5.39 0.12 0.85 3.8 4.19 1.45 0.16 3.88 95.86 TOR0000 RG Channel Lake Member 64.69 0.79 14.38 5.76 0.13 0.97 3.86 4.24 1.61 0.18 3.12 96.61 Major elements concentrations were analyzed at Michigan State University on a Bruker S4 Pioneer X - Ray Fluorescence Spectrometer (XRF). 89 Table 1 ( con ) Sample Unit Member SiO2 (%) TiO2 (%) Al2O3 (%) Fe2O3 (%) MnO (%) MgO (%) CaO (%) Na2O (%) K2O (%) P2O5 (%) LOI (%) Sum (%) TOR0000 RH Channel Lake Member 64.56 0.8 14.32 5.59 0.14 0 .98 3.58 4.53 1.9 0.18 3.17 96.58 TOR0000 RI Channel Lake Member 67.36 0.65 14.19 4.97 0.1 1.45 0.63 4.45 3.96 0.14 1.83 97.9 TOR0000 RJ Channel Lake Member 58.33 1.51 15.03 8.76 0.12 3.64 4.29 3.55 2.27 0.44 1.89 97.94 TOR0000 RK Channel Lake Member 57.95 1.41 14.85 9.11 0.16 3.74 2.03 3.16 4.22 0.41 2.77 97.04 TOR0000 RH Channel Lake Member 64.56 0.8 14.32 5.59 0.14 0.98 3.58 4.53 1.9 0.18 3.17 96.58 TOR0000 RI Channel Lake Member 67.36 0.65 14.19 4.97 0.1 1.45 0.63 4.45 3.96 0.14 1.83 97.9 TOR0000 RJ Cha nnel Lake Member 58.33 1.51 15.03 8.76 0.12 3.64 4.29 3.55 2.27 0.44 1.89 97.94 TOR0000 RL Channel Lake Member 59.02 1.4 15.08 8.41 0.13 2.92 4.37 3.78 2.02 0.42 2.29 97.55 TOR0000 RM Channel Lake Member 62.93 1.12 14.36 7.17 0.11 2.49 1.01 4.46 3.59 0.29 2.25 97.53 TOR0000 RN Channel Lake Member 63.68 1.09 13.88 7.17 0.12 2.97 0.88 3.05 4.59 0.22 2.14 97.65 TOR0000 RO Channel Lake Member 61.95 1.1 14.61 7.19 0.16 3.68 0.78 3. 69 4.27 0.22 2.13 97.65 TOR0000 T1 Channel Lake Member 61.63 1.24 14.73 7.3 0.14 3 .47 1.44 2.82 4.4 0.31 2.34 97.48 TOR0000 S8 Channel Lake Member 58.3 1.74 15.35 9.24 0.15 3.43 1.71 7.15 0.08 0.44 2.3 97.59 TOR0000 S5 Quebec Harbor Member altered 68.51 0. 66 13.51 5.01 0.06 1.13 0.42 4.12 4.49 0.13 1.72 98.04 TOR0000 S6 Quebec Harbor Me mber altered 48.9 2.1 16.74 12.78 0.2 5.78 2.25 5.26 1.55 0.53 3.71 96.09 TOR0000 SN Quebec Harbor Member altered 65.29 1.5 11.49 7.34 0.09 2.7 2.9 5.48 0.09 0.31 2.73 97.19 TOR0000 SJ Quebec Harbor Member unaltered 65.98 0.81 13.6 4.92 0.12 0.79 3.53 4.23 1.36 0.16 4.27 95.5 TOR0000 SK Quebec Harbor Member unaltered 66.45 0.78 13.63 4.69 0.11 0.68 3.22 4.4 1.93 0.15 3.71 96.04 TOR0000 SM South Shore Member 52.94 2.55 14.13 12 .27 0.33 3.57 8.19 2.72 0.45 0.53 2.12 97.68 TOR0000 SL South Shore Member 52.74 2 .47 14.05 12.62 0.22 3.54 8.33 2.72 0.41 0.64 2.08 97.74 TOR0000 SI South Shore Member 54.8 2.44 15.86 9.11 0.1 2.05 4.44 5.34 3.01 0.41 2.23 97.56 TOR0000 SH South Shore Mem ber 51.1 2.06 13.32 14.16 0.39 5.01 6.93 3.04 1.31 0.29 2.23 97.61 TOR0000 S4 Sout h Shore Member 54.77 2.2 14.01 11.07 0.15 4.23 5.09 4.12 1.82 0.32 2.05 97.78 TOR0000 S3 South Shore Member 51.86 2.14 13.51 14.55 0.23 4.37 8.12 2.95 1.13 0.33 0.7 99.19 TO R0000 S2 South Shore Member 50.3 2.38 13.29 15.12 0.23 4.44 9.1 2.77 0.41 0.27 1.54 98.31 TOR0000 S1 South Shore Member 47.49 2.46 14.05 15.82 0.32 5.23 9.47 2.57 0.38 0.28 1.8 98.07 TOR0000 S0 South Shore Member 53.1 2.39 13.59 12.65 0.23 3.64 7.93 3.23 0. 57 0.67 1.86 98 TOR0000 RZ South Shore Member 50.5 2.48 14.3 12.86 0.45 4.38 8.52 2.84 0.44 0.69 2.23 97.46 TOR0000 RY South Shore Member 49.85 2.49 14.45 13.17 0.31 3.62 8.53 2.67 0.41 0.77 3.6 96.27 TOR0000 RX South Shore Member 52.47 2.52 13.74 14.27 0. 17 3.49 5.91 3.76 1.58 0.38 1.58 98.29 TOR0000 RW South Shore Member 51.7 2.45 13. 82 13.65 0.27 4.18 7.81 2.88 1.26 0.39 1.49 98.41 TOR0000 RV South Shore Member 53.4 2.44 14.04 11.49 0.25 4.57 3.83 4.41 1.9 0.66 2.81 96.99 TOR0000 RU South Shore Member 53 .77 2.54 14.19 11.52 0.25 4.74 3.14 4.22 2.35 0.56 2.49 97.28 TOR0000 RT South Sho re Member 53.94 2.41 13.78 12.55 0.26 3.65 7.21 3.04 1.6 0.56 0.88 99 TOR0000 RS South Shore Member 52.99 2.43 13.76 12.97 0.28 3.77 7.77 3.2 0.94 0.52 1.26 98.63 TOR0000 RR South Shore Member 51.57 2.44 14.06 12.98 0.26 4.04 7.71 2.96 1.4 0.72 1.74 98.14 TOR0000 RQ South Shore Member 52.27 2.59 13.64 13.19 0.28 3.79 7.98 3.27 1.19 0.43 1.25 98.63 TOR0000 RP Davieaux Island Member 70.9 0.19 13.75 2.43 0.04 0.78 0.15 1.16 8.62 0.02 1.75 98.04 90 Table 2 : Trace Elements Concentrations Sample Unit Member Analy sis Day Sc (ppm) V (ppm) Cr (ppm) Co (ppm) Ni (ppm) Ga (ppm) Rb (ppm) Sr (ppm) TOR0000 SB Michipicoten Island Intrusive 3 15.3 91 2.71 14.2 5.92 23.2 104.7 199 TOR0000 SD Cuesta Lower Flow 3 25.1 155 11.64 23.5 14.85 22.4 80.7 176 TOR0000 SE Cuesta Lower F low 3 27.9 161 12.18 25.2 16.24 22.0 66.3 165 TOR0000 SF Cuesta Lower Flow 3 26.4 160 12.43 24.4 15.35 21.4 67.7 189 TOR0000 SG Cuesta Lower Flow 3 26.1 157 12.27 23.0 14.74 22.8 73.6 157 TOR0000 QQ Cuesta Upper Flow 1 20.3 82 5.06 16.0 7.04 23.4 118.8 123 TOR0000 QR Cuesta Upper Flow 1 19.4 78 4.82 15.2 6.84 25.5 41.6 137 TOR0000 QS Cuesta Upper Flow 3 18.6 63 4.21 13.6 5.25 23.8 58.2 612 TOR0000 QJ Channel Lake Member 1 10.1 7 BDL 3.4 BDL 22.2 88.1 454 TOR0000 QK Channel Lake Member 1 10.5 7 BDL 3.2 BDL 2 2.3 78.4 325 TOR0000 QL Channel Lake Member 1 10.1 7 BDL 3.2 BDL 21.1 161.3 81 TOR0000 QM Channel Lake Member 1 11.0 11 BDL 4.2 BDL 22.2 81.6 232 TOR0000 QN Channel Lake Member 1 10.4 14 BDL 2.8 BDL 20.5 144.1 89 TOR0000 QO Channel Lake Member 1 10.7 11 BD L 3.8 BDL 21.7 79.9 292 TOR0000 QT Ch annel Lake Member 1 11.7 17 BDL 4.8 0.92 22.5 90.8 270 TOR0000 QU Channel Lake Member 1 11.3 14 BDL 4.3 1.04 22.6 82.8 330 TOR0000 QV Channel Lake Member 1 11.1 14 BDL 4.2 0.97 21.6 98.6 283 TOR0000 QW Channel Lake Memb er 1 11.7 21 BDL 4.9 1.02 22.9 94.6 3 22 TOR0000 QX Channel Lake Member 1 10.5 7 BDL 2.2 BDL 20.6 154.9 58 TOR0000 QY Channel Lake Member 1 13.3 38 BDL 6.9 2.05 23.1 88.5 319 TOR0000 QZ Channel Lake Member 1 12.5 25 BDL 5.6 1.09 22.0 86.9 369 TOR0000 R1 Cha nnel Lake Member 2 21.6 136 3.02 20.3 13.01 20.9 123.7 145 TOR0000 R3 Channel Lake Member 2 24.3 153 4.11 20.8 12.75 23.0 37.9 630 TOR0000 R4 Channel Lake Member 2 23.8 161 4.12 22.4 11.79 23.1 41.2 412 TOR0000 R5 Channel Lake Member 2 22.9 160 3.73 16.5 1 0.73 20.8 95.7 159 TOR0000 R6 Channel Lake Member 2 16.1 39 1.55 8.3 2.73 23.1 54.0 728 TOR0000 R7 Channel Lake Member 2 18.0 54 1.90 9.5 3.18 22.6 53.3 733 Trace elements analyses using a Photon - Machines Analyte G2 Excimer laser and Thermo Scientific ICA P Q quadrupole inductively coupled plasma mass spectrometer (ICP - MS). Deviation on replicated a nalyses is less than 5% except for low concentrations (< 2 p p m) of Cr and Ni . Below detection limit (BLD) is 3 times the gas blank. 91 Table 2 (c ) Sample U nit Member A nalysis Day Sc (ppm) V (ppm) Cr (ppm) Co (ppm) Ni (ppm) Ga (ppm) Rb (ppm) Sr (ppm) TOR0000 R8 Channel Lake Member 3 27.3 126 BDL 18.8 1.19 21.1 31.3 509 TOR0000 R9 Channel Lake Member 2 26.5 134 BDL 19.2 1.25 22.9 30.9 592 TOR0000 RA Channel La ke Member 1 26.1 125 BDL 18.5 1.62 22.2 30.8 569 TOR0000 RB Channel Lake Member 2 28.6 148 BDL 21.0 1.07 22.7 35.0 175 TOR0000 RC Channel Lake Member 2 26.8 138 0.97 20.1 1.72 22.9 28.2 515 TOR0000 RD Channel Lake Member 2 27.7 139 BDL 20.8 1.57 23.3 28.4 522 TOR0000 RE Channel Lake Member 2 27.3 146 0.93 19.5 2.04 23.4 32.3 576 TOR0000 RF Channel Lake Member 2 13.6 15 1.06 5.1 1.38 23.3 57.1 564 TOR0000 RG Channel Lake Member 2 14.9 22 1.07 6.3 1.63 23.5 60.0 495 TOR0000 RH Channel Lake Member 2 15.5 21 1. 17 6.3 1.65 23.1 57.0 395 TOR0000 RI Channel Lake Member 3 12.6 8 BDL 4.5 1.07 21.0 149.9 154 TOR0000 RJ Ch annel Lake Member 1 22.1 131 4.50 19.0 16.72 23.3 63.5 184 TOR0000 RK Channel Lake Member 1 20.5 111 4.36 17.4 15.49 22.5 166.7 219 TOR0000 RL Channe l Lake Member 2 21.6 121 4.56 19.4 17.82 22.1 64.8 268 TOR0000 RM Channel Lake Member 2 20.1 62 1.50 10.8 3 .71 21.6 108.9 235 TOR0000 RN Channel Lake Member 2 21.7 103 1.40 13.0 5.37 22.8 113.4 224 TOR0000 T1 Channel Lake Member 4 18.9 74 2.00 13.4 5.04 - 94.2 140 TOR0000 S5 Quebec Harbor Member altered 3 11.2 8 BDL 3.0 0.62 18.8 142.1 135 TOR0000 S6 Quebec Har bor Member altered 3 31.5 215 6.16 26.5 18.27 23.8 37.6 154 TOR0000 SN Quebec Harbor Member altered 4 22.8 148 3.14 16.3 9.33 - 2.0 30 TOR0000 SJ Que bec Harbor Member unaltered 4 12.8 15 BDL 5.2 0.91 - 81.0 209 TOR0000 SK Quebec Harbor Member 4 12.4 14 BD L 5.0 0.81 - 113.6 325 TOR0000 SM South Shore Member 3 33.4 182 BDL 23.4 0.70 20.1 7.7 1158 TOR0000 SL South Shore Member 3 32.9 177 BDL 22.6 2.27 22 .6 7.3 774 TOR0000 SI South Shore Member 3 44.2 315 22.68 20.1 17.21 14.9 114.6 264 TOR0000 SH South Shore Member 3 40.1 382 9.24 41.7 21.19 22.3 25.4 191 TOR0000 S4 South Shore Member 3 29.5 253 7.98 36.3 21.32 16.3 51.2 177 TOR0000 S3 South Shore Member 3 39.6 361 BDL 42.4 19.07 19.4 34.1 174 TOR0000 S2 South Shore Member 3 41.4 440 10.92 43.7 34.59 20.1 12.0 301 TOR0000 S1 South Shore Member 3 42.9 461 11.46 45.6 37.33 21.5 4.9 332 TOR0000 S0 South Shore Member 3 32.3 192 BDL 26.8 3.01 20.9 19.4 459 TOR 0000 RZ South Shore Member 3 32.3 208 1.27 27.6 3.52 21.1 7.7 2104 92 Table 2 Sample Unit Member Analysis Day Sc (ppm) V (ppm) Cr (ppm) Co (ppm) Ni (ppm) Ga (ppm) Rb (ppm) Sr (ppm) TOR0000 RY South Shore Member 3 33.0 196 BDL 26 .9 4.32 21.3 6.4 457 TOR0000 RX South Shore Member 2 37.8 329 1.10 34.1 4.26 21.1 75.9 235 TOR0000 RW South Shore Member 2 36.2 308 BDL 35.6 3.70 23.2 40.7 172 TOR0000 RV South Shore Member 2 31.5 194 1.02 28.6 2.61 23.3 57.2 396 TOR0000 RU South Shore Mem ber 2 35.5 213 0.91 27.6 1.64 21.7 76.9 409 TOR0000 RT South Shore Member 1 32.6 19 8 BDL 27.8 1.18 20.9 43.7 190 TOR0000 RS South Shore Member 1 34.3 211 BDL 29.7 1.10 21.6 20.3 285 TOR0000 RR South Shore Member 1 33.0 215 1.40 26.8 4.19 23.9 32.8 190 TOR 0000 RQ South Shore Member 2 37.0 262 1.99 31.1 7.90 23.3 38.1 185 TOR0000 RP Daviea ux Island Member 3 3.0 6 BDL 2.4 2.52 20.6 243.2 42 93 Table 2 ( cont d ) Sample Y (ppm) Zr (ppm) Nb (ppm) Cs (ppm) Ba (ppm) La (ppm) Ce (ppm) Pr (ppm) Nd (ppm) TOR0000SB 93 .2 469 46.1 81.41 756 60.5 123.3 14.5 57.6 TOR0000SD 71.5 468 34.6 1.65 689 45.9 9 6.5 11.8 49.3 TOR0000SE 78.6 493 35.7 1.03 657 47.9 98.5 12.3 52.1 TOR0000SF 73.4 465 34.5 0.98 613 46.1 95.6 11.8 49.7 TOR0000SG 70.4 467 34.8 1.03 600 47.2 97.6 11.9 50 .0 TOR0000QQ 84.8 563 43.4 3.40 1939 58.5 125.2 14.8 58.9 TOR0000QR 81.1 564 41.7 3.52 451 76.6 152.3 17.4 65.9 TOR0000QS 81.1 577 40.4 51.96 759 55.4 112.8 13.8 57.3 TOR0000QJ 76.9 688 50.2 220.48 1035 63.9 131.0 14.8 55.1 TOR0000QK 77.6 699 51.4 47. 45 1008 64.7 132.1 15.1 55.9 TOR0000QL 70.9 679 47.8 1.52 774 55.9 116.0 12.9 47.4 TOR0000QM 76.9 693 50.6 34.41 983 64.3 130.0 14.9 55.7 TOR0000QN 66.9 659 46.8 1.41 870 61.4 121.3 13.8 51.4 TOR0000QO 75.6 689 50.5 34.93 1022 64.4 131.5 15.1 55.9 TOR 0000QT 78.3 695 49.6 26.97 998 64.2 131.4 15.2 56.2 TOR0000QU 76.4 694 49.1 33.37 964 63.5 127.4 14.7 55.2 TOR0000QV 72.2 647 48.5 28.66 953 60.8 125.3 14.1 52.8 TOR0000QW 74.9 666 49.3 34.18 967 62.5 129.2 14.7 54.7 TOR0000QX 64.1 675 48.3 2.19 882 60 .2 120.7 13.9 51.7 TOR0000QY 71.0 626 46.7 34.75 941 58.0 120.4 13.9 51.5 TOR0000 QZ 75.8 659 47.6 37.48 943 61.6 124.6 14.2 53.7 TOR0000R1 66.3 535 39.0 5.74 860 45.6 104.3 12.4 50.7 TOR0000R3 62.4 470 33.2 77.61 579 44.5 92.0 11.3 46.0 TOR0000R4 61.5 458 32.9 149.20 619 41.9 90.1 10.8 44.8 TOR0000R5 55.4 428 31.5 3.40 1156 39.6 88 .2 10.6 42.9 TOR0000R6 72.9 706 47.1 161.88 821 54.6 120.5 14.0 54.6 TOR0000R7 75.7 704 44.7 131.86 787 54.8 115.2 13.7 54.7 TOR0000R8 76.0 442 33.8 59.01 496 44.4 93.2 1 1.7 49.5 TOR0000R9 71.4 417 33.7 83.61 514 41.6 90.9 11.5 48.6 TOR0000RA 72.0 418 32.8 63.53 499 42.8 88.3 11.1 46.0 TOR0000RB 78.8 444 35.3 4.66 393 39.5 94.5 12.1 50.9 TOR0000RC 72.4 418 32.7 72.67 505 40.8 88.7 11.2 47.4 TOR0000RD 72.2 414 32.5 66. 80 496 40.4 87.2 11.0 47.0 TOR0000RE 71.8 428 33.0 70.84 466 40.4 86.5 10.8 46.5 TOR0000RF 79.7 828 49.5 117.29 898 59.8 126.2 14.8 57.6 TOR0000RG 79.5 826 50.0 96.35 903 59.7 127.6 14.8 57.9 TOR0000RH 83.3 851 50.3 81.54 974 61.2 125.9 15.1 59.2 TOR0 000RI 79.6 866 49.1 1.85 1174 98.8 163.4 18.1 66.0 TOR0000RJ 58.7 460 32.5 1.70 59 3 48.0 94.0 11.2 44.1 94 Table 2 (c ) Sample Y (ppm) Zr (ppm) Nb (ppm) Cs (ppm) Ba (ppm) La (ppm) Ce (ppm) Pr (ppm) Nd (ppm) TOR0000RK 59.0 483 33.8 18.83 655 53.4 98 .7 11.6 46.1 TOR0000RL 58.8 481 34.4 2.40 567 46.5 98.8 11.9 47.5 TOR0000RM 84.1 763 48.0 4.54 923 51.8 110.7 13.4 54.3 TOR0000RN 75.4 769 48.3 6.15 795 49.3 112.0 13.2 52.5 TOR0000T1 71.1 640 42.3 4.09 618 49.0 105.6 12.5 49.6 TOR0000S5 75.4 659 49.3 2.53 1025 57.1 117.8 13.0 49.9 TOR0000S6 71.7 511 37.6 3.46 480 44.2 104.5 12.8 52.5 TOR0000SN 47.7 344 24.8 0.28 39 27.1 57.7 6.8 28.4 TOR0000SJ 78.7 680 50.0 47.14 883 61.2 124.8 14.7 57.4 TOR0000SK 79.6 691 50.6 18.92 900 62.0 126.4 14.8 57.1 TOR0 000SM 63.4 285 25.1 0.45 354 31.4 67.8 8.7 37.6 TOR0000SL 63.0 285 25.8 0.42 289 35.7 74.4 9.5 40.4 TOR0000SI 62.6 360 28.5 24.89 775 37.1 80.4 10.0 41.7 TOR0000SH 49.9 262 20.5 0.50 449 28.7 60.2 7.4 31.1 TOR0000S4 32.9 195 15.5 3.00 464 20.0 43.4 5.3 22.3 TOR0000S3 49.9 242 19.3 0.59 343 26.8 56.4 7.1 29.7 TOR0000S2 50.2 208 18. 3 6.97 257 22.8 47.9 6.3 27.1 TOR0000S1 51.1 210 18.7 0.51 233 23.7 50.8 6.5 28.2 TOR0000S0 63.1 279 24.5 15.75 354 33.9 71.1 9.2 40.0 TOR0000RZ 63.1 280 25.5 1.52 409 34. 9 77.3 9.7 41.1 TOR0000RY 67.7 296 26.7 0.48 290 37.4 80.8 10.2 43.9 TOR0000RX 5 6.8 263 22.0 8.34 495 27.4 58.5 7.6 33.5 TOR0000RW 55.5 250 21.1 0.49 344 26.8 58.1 7.5 32.7 TOR0000RV 61.2 282 26.2 2.32 866 31.6 73.2 9.4 40.3 TOR0000RU 61.9 303 26.8 1. 79 1180 33.3 73.2 9.5 41.0 TOR0000RT 59.4 271 24.2 1.27 405 32.3 67.9 8.7 36.4 T OR0000RS 58.8 274 23.9 4.58 362 31.9 66.4 8.5 36.0 TOR0000RR 64.3 282 25.2 0.64 391 35.3 73.3 9.4 39.8 TOR0000RQ 60.0 263 22.6 0.59 362 28.5 62.3 8.1 35.7 TOR0000RP 108.2 270 51.0 16.94 1180 42.5 88.3 9.1 32.3 95 Table 2 ( cont d ) Sample Sm (ppm) Eu (pp m) Gd (ppm) Tb (ppm) Dy (ppm) Ho (ppm) Er (ppm) Tm (ppm) Yb (ppm) TOR0000SB 13.5 2.52 14.5 2.44 15.5 3.37 9.74 1.56 10.22 TOR0000SD 11.5 3.51 12.3 1.98 12.4 2.65 7.52 1.16 7.62 TOR0000SE 12.3 3.64 13.3 2.10 13.2 2.85 7.96 1.25 8.03 TOR0000SF 11.8 3.55 12.6 2.00 12.4 2.66 7.59 1.18 7.75 TOR0000SG 11.8 3.63 12.3 2.02 12.5 2.62 7.35 1.15 7.43 TOR0000QQ 13.6 3.70 14.4 2.39 15.0 3.10 9.00 1.41 8.70 TOR0000QR 12.9 3.50 13.3 2 .15 13.7 2.95 8.87 1.35 8.91 TOR0000QS 13.1 3.70 13.9 2.21 13.9 2.96 8.61 1.32 8. 50 TOR0000QJ 12.1 2.26 12.1 2.04 13.0 2.78 8.33 1.30 8.59 TOR0000QK 12.2 2.32 12.2 2.09 13.1 2.80 8.42 1.35 8.69 TOR0000QL 10.6 2.32 11.1 1.91 12.4 2.66 8.00 1.25 8.40 TO R0000QM 12.1 2.33 12.4 2.09 13.2 2.81 8.48 1.35 8.85 TOR0000QN 11.2 2.12 11.1 1.8 7 11.7 2.47 7.29 1.16 7.47 TOR0000QO 12.3 2.36 12.3 2.06 13.1 2.81 8.43 1.31 8.81 TOR0000QT 12.5 2.43 12.5 2.10 13.3 2.84 8.52 1.33 8.81 TOR0000QU 12.0 2.36 12.2 2.07 13.2 2.80 8.33 1.33 8.57 TOR0000QV 11.5 2.27 11.5 1.94 12.4 2.59 7.75 1.26 8.06 TOR0 000QW 12.1 2.41 12.0 2.03 12.8 2.71 8.16 1.26 8.43 TOR0000QX 11.1 2.11 11.0 1.85 11.7 2.46 7.40 1.21 7.87 TOR0000QY 11.3 2.32 11.4 1.94 12.3 2.59 7.70 1.25 7.89 TOR0000QZ 11.9 2.43 12.1 2.03 12.9 2.71 8.13 1.31 8.33 TOR0000R1 11.4 2.88 11.8 1.88 11.4 2 .36 6.84 1.02 6.99 TOR0000R3 10.5 2.70 10.9 1.73 10.9 2.28 6.55 0.98 6.63 TOR0000R4 10.1 2.67 10.6 1.67 10.7 2.18 6.38 0.95 6.53 TOR0000R5 9.7 2.60 9.9 1.58 9.9 2.04 5.87 0.88 5.94 TOR0000R6 12.1 2.86 12.3 1.98 12.6 2.62 7.75 1.19 8.25 TOR0000R7 12.1 2.88 12.5 2.03 12.7 2.66 7.89 1.21 8.18 TOR0000R8 11.7 3.19 12.9 2.13 13.0 2.78 7.99 1.20 7.75 TOR0000R9 11.8 3.16 12.7 2.01 12.7 2.63 7.54 1.11 7.48 TOR0000RA 11.1 3.10 1 2.3 2.07 12.7 2.67 7.66 1.20 7.33 TOR0000RB 12.4 3.02 13.4 2.17 13.5 2.77 8.15 1. 19 7.92 TOR0000RC 11.6 3.05 12.7 1.97 12.6 2.62 7.56 1.11 7.47 TOR0000RD 11.2 3.06 12.4 1.98 12.5 2.59 7.51 1.13 7.44 TOR0000RE 10.9 3.01 12.1 1.96 12.1 2.53 7.30 1.09 7.2 6 TOR0000RF 12.9 2.84 12.8 2.09 13.3 2.80 8.43 1.29 8.94 TOR0000RG 12.6 2.86 12. 6 2.09 13.2 2.78 8.34 1.29 8.82 TOR0000RH 12.9 2.89 13.3 2.16 13.8 2.90 8.73 1.35 9.25 TOR0000RI 12.2 2.60 12.0 2.05 13.1 2.87 8.57 1.39 9.00 TOR0000RJ 9.9 2.74 10.4 1.68 10.3 2.14 6.14 0.96 5.94 96 Table 2 (c ) Sample Sm (ppm) Eu (ppm) Gd (ppm) T b (ppm) Dy (ppm) Ho (ppm) Er (ppm) Tm (ppm) Yb (ppm) TOR0000RK 10.8 3.04 10.7 1.70 10.4 2.14 6.21 0.97 6.06 TOR0000RL 10.3 2.61 10.7 1.67 10.3 2.09 6.13 0.91 6.14 TOR0000R M 12.5 2.96 13.5 2.22 14.4 3.01 9.07 1.37 9.43 TOR0000RN 12.1 2.65 12.2 2.05 13.3 2.80 8.25 1.27 8.62 TOR0000T1 10.9 2.71 11.7 1.91 12.1 2.59 7.58 1.16 7.68 TOR0000S5 10.8 2.20 11.1 1.94 12.5 2.72 8.02 1.26 8.32 TOR0000S6 12.0 2.94 12.5 1.96 12.3 2.62 7.51 1.17 7.48 TOR0000SN 7.1 2.25 8.0 1.33 8.3 1.75 5.07 0.77 5.02 TOR0000SJ 12. 1 2.49 12.5 2.05 12.9 2.76 8.21 1.28 8.42 TOR0000SK 12.0 2.43 12.4 2.05 13.0 2.79 8.20 1.28 8.54 TOR0000SM 9.4 2.78 10.6 1.73 10.9 2.35 6.55 1.01 6.44 TOR0000SL 10.1 2.97 11.0 1.77 10.8 2.32 6.50 1.01 6.44 TOR0000SI 9.8 2.66 10.3 1.66 10.4 2.27 6.45 0. 99 6.37 TOR0000SH 7.5 2.19 8.2 1.34 8.4 1.81 5.12 0.79 5.13 TOR0000S4 5.3 1.64 5.8 0.93 5.9 1.23 3.40 0.52 3.34 TOR0000S3 7.3 2.19 8.3 1.35 8.6 1.81 5.21 0.79 5.20 TOR000 0S2 7.0 2.07 8.1 1.34 8.5 1.83 5.20 0.80 5.17 TOR0000S1 7.3 2.21 8.2 1.37 8.7 1.8 7 5.31 0.81 5.28 TOR0000S0 9.8 2.83 11.0 1.76 10.8 2.29 6.44 0.97 6.17 TOR0000RZ 10.1 3.01 11.0 1.78 10.9 2.31 6.53 0.97 6.19 TOR0000RY 10.7 3.17 12.0 1.91 11.7 2.47 6.97 1.03 6.50 TOR0000RX 8.5 2.64 9.8 1.58 10.0 2.07 5.81 0.86 5.72 TOR0000RW 8.4 2.5 1 9.6 1.54 9.8 2.04 5.85 0.87 5.74 TOR0000RV 10.1 2.99 11.1 1.75 11.0 2.30 6.45 0.94 6.33 TOR0000RU 10.1 2.97 11.4 1.78 11.4 2.33 6.53 0.99 6.65 TOR0000RT 9.2 2.70 10.3 1. 70 10.5 2.19 6.27 0.98 6.02 TOR0000RS 9.0 2.76 10.3 1.69 10.6 2.19 6.30 0.98 5.99 TOR0000RR 9.9 2.96 11.3 1.82 11.4 2.36 6.68 1.04 6.25 TOR0000RQ 9.1 2.77 10.5 1.65 10.6 2.17 6.20 0.91 6.07 TOR0000RP 7.7 0.86 11.1 2.38 16.4 3.63 10.70 1.65 10.30 97 Tab le 2 ( cont d ) Sample Lu (ppm) Hf (ppm) Ta (ppm) Pb (ppm) Th (ppm) U (ppm) TOR0000 SB 1.50 12.2 3.21 18.0 14.97 3.75 TOR0000SD 1.17 10.8 2.36 11.7 9.72 2.68 TOR0000SE 1.25 11.4 2.50 10.5 10.19 2.70 TOR0000SF 1.18 10.8 2.35 10.7 9.65 2.59 TOR0000SG 1.14 10.8 2.37 12.9 9.64 2.72 TOR0000QQ 1.35 13.3 2.95 5.4 11.90 3.35 TOR0000QR 1.37 13.4 2.89 6.8 12.10 3.58 TOR0000QS 1.30 13.5 2.85 14.7 12.30 3.35 TOR0000QJ 1.34 16.5 3.41 18.7 17.36 4.76 TOR0000QK 1.36 16.5 3.45 17.6 17.24 4.42 TOR0000QL 1.30 16.2 3. 31 6.8 16.92 4.53 TOR0000QM 1.35 16.6 3.48 17.9 17.14 4.28 TOR0000QN 1.17 15.6 3 .20 13.2 15.87 4.13 TOR0000QO 1.33 16.6 3.44 17.4 17.31 4.51 TOR0000QT 1.35 16.6 3.44 17.7 17.18 4.54 TOR0000QU 1.35 16.5 3.36 17.2 16.71 4.34 TOR0000QV 1.26 15.4 3.24 17 .3 15.97 4.46 TOR0000QW 1.28 15.8 3.32 18.3 16.38 4.39 TOR0000QX 1.22 16.1 3.37 16.0 16.60 4.20 TOR0000QY 1.24 15.0 3.11 17.7 15.28 4.39 TOR0000QZ 1.32 15.8 3.22 16.2 15.83 4.12 TOR0000R1 1.02 12.0 2.58 6.0 10.47 2.64 TOR0000R3 0.97 10.8 2.24 12.5 9. 30 2.22 TOR0000R4 0.95 10.5 2.17 9.9 9.09 2.39 TOR0000R5 0.88 9.8 2.07 8.5 8.53 2.16 TOR0000R6 1.21 15.6 3.08 14.0 13.50 3.46 TOR0000R7 1.24 15.7 2.92 13.2 13.17 3.46 TOR0000R8 1.16 10.5 2.32 9.6 8.97 2.29 TOR0000R9 1.08 10.0 2.30 9.8 8.65 2.15 TOR0 000RA 1.15 10.4 2.25 13.0 8.49 2.15 TOR0000RB 1.15 10.7 2.45 6.5 9.07 2.26 TOR00 00RC 1.11 10.1 2.26 8.9 8.57 2.13 TOR0000RD 1.08 9.9 2.20 8.7 8.37 2.14 TOR0000RE 1.09 10.1 2.21 8.5 8.53 2.24 TOR0000RF 1.32 18.4 3.30 15.8 15.67 4.16 TOR0000RG 1.33 18. 0 3.29 15.6 15.26 4.09 TOR0000RH 1.38 19.0 3.37 26.7 16.00 4.02 TOR0000RI 1.43 1 8.9 3.36 4.8 16.12 4.54 TOR0000RJ 0.93 10.7 2.20 9.8 9.34 2.38 98 Table 2 (c ) Sample Lu (ppm) Hf (ppm) Ta (ppm) Pb (ppm) Th (ppm) U (ppm) TOR0000RK 0.96 11.3 2.27 11 .1 9.71 2.42 TOR0000RL 0.92 10.9 2.25 10.6 9.87 2.46 TOR0000RM 1.40 17.3 3.32 4. 6 14.26 3.07 TOR0000RN 1.27 17.0 3.24 5.8 14.40 3.38 TOR0000T1 1.16 14.7 2.84 5.3 12.47 3.36 TOR0000S5 1.27 15.2 3.32 12.7 15.70 4.12 TOR0000S6 1.16 11.7 2.50 5.0 10.34 2 .78 TOR0000SN 0.74 8.1 1.65 4.7 6.83 1.78 TOR0000SJ 1.29 16.1 3.40 16.5 16.68 4. 46 TOR0000SK 1.31 16.4 3.41 16.7 17.05 4.59 TOR0000SM 0.98 7.2 1.79 8.0 6.71 1.99 TOR0000SL 0.97 7.1 1.84 5.9 6.71 1.89 TOR0000SI 0.97 8.6 1.81 8.8 7.21 3.22 TOR0000SH 0 .78 6.3 1.33 9.2 5.44 1.35 TOR0000S4 0.50 4.7 1.01 10.7 3.98 1.30 TOR0000S3 0.77 5.9 1.30 4.8 5.08 1.22 TOR0000S2 0.78 5.4 1.26 4.9 4.58 1.18 TOR0000S1 0.78 5.4 1.31 2.9 4.65 1.21 TOR0000S0 0.94 7.0 1.76 6.9 6.52 1.78 TOR0000RZ 0.93 7.1 1.81 9.1 6.69 1.93 TOR0000RY 0.98 7.4 1.92 7.3 7.03 1.97 TOR0000RX 0.84 6.7 1.62 8.3 5.77 1.5 6 TOR0000RW 0.86 6.4 1.55 6.9 5.51 1.53 TOR0000RV 0.91 7.2 1.89 11.6 6.78 1.87 TOR0000RU 0.98 7.8 2.00 10.9 7.29 1.87 TOR0000RT 0.95 7.0 1.73 4.5 6.35 1.81 TOR0000RS 0.9 3 7.1 1.75 6.6 6.35 1.70 TOR0000RR 0.98 7.3 1.82 6.8 6.61 1.89 TOR0000RQ 0.89 6. 8 1.63 6.0 5.81 1.46 TOR0000RP 1.53 9.4 4.27 11.1 26.51 2.92 99 Table 3 : Measured Isotopic Values Name Unit 176 Hf/ 177 Hf 2se 143 Nd/ 144 Nd 2se 87 Sr/ 86 Sr 2se TOR0000QR Cuesta Upper Flow 0. 282470 6E - 06 0.512246 4E - 06 0.713304 5E - 06 TOR0000QW Channel Lake Member 0.282453 8E - 06 0.512161 3E - 06 0.720782 5E - 06 TOR0000RJ Channel Lake Member 0.282433 5E - 06 0.512344 5E - 06 0.720515 5E - 06 TOR0000RP Davieaux Island Member 0.282576 4E - 06 0.512387 2.1 E - 05 0.918183 7E - 06 TOR0000RQ South Shore Member 0.282577 8E - 06 0.512672 8E - 06 0.713099 5E - 06 TOR0000S1 South Shore Member 0.282173 6E - 06 0.512139 3E - 06 0.729279 6E - 06 TOR0000SD Cuesta Lower Flow 0.282459 4E - 06 0.512503 5E - 06 0.727211 6E - 06 TOR0000SJ Q uebec Harbor Member unaltered 0.282416 9E - 06 0.512143 4E - 06 0.720486 5E - 06 Standards BCR - 2 0.282880 6E - 06 0.512633 6E - 06 0.705006 8E - 06 duplicate 0.512641 6E - 06 0.705005 6E - 06 BIR - 1 0.513097 8E - 06 duplic ate K1919 (generic BHVO) 0.512992 5E - 06 JMC 475 0.282155 JMC 475 0.282157 4E - 06 JMC 475 0.282157 8E - 06 JMC 475 0.282157 4E - 06 JMC 475 0.282161 4E - 06 JMC 475 0.282152 5E - 06 JMC 475 0.282154 4E - 06 JM C 475 0.282148 3E - 06 Samples were measured on a Nu Plasma HR multi - collector inductively coupled plasma mass spectrometer (MC - ICPMS). Uncertainties are reported in terms of standard error (se ) 100 Table 3 (c ) Name Unit 208Pb /204Pb 1se 207 Pb/ 204 Pb 1se 206 Pb/ 204 Pb 1se TOR0000QR Cuesta Upper Flow 41.0815 0.0011 15.8271 0.0003 20.6698 0.0003 TOR0000QW Channel Lake Member 40.3183 0.0009 15.7396 0.0006 20.0029 0.0004 TOR0000RJ Channel Lake Member 39.4122 0.0005 15.6832 0.0002 19 .2491 0.0002 TOR0000RP Davieaux Island Member 42.5978 0.000 8 15.7616 0.0003 20.9796 0.0003 TOR0000RQ South Shore Member 39.4615 0.0007 15.7054 0.0003 19.4941 0.0003 TOR0000S1 South Shore Member 43.6013 0.0008 15.9285 0.0003 22.8858 0.0003 TOR0000SD Cue sta Lower Flow 38.6391 0.0004 15.6552 0.0002 18.7893 0.0002 TOR0000SJ Q uebec Harbor Member unaltered 43.9524 0.0010 15.9487 0.0004 23.1900 0.0005 Standards BCR - 2 38.6569 0.0008 15.5999 0. 0003 18.7375 0.0003 duplicate BIR - 1 3 8.4700 0.0007 15.6535 0.0002 18.8480 0.0003 duplicate 38.4819 0.0006 15.6530 0.0003 18.8518 0.0003 101 Table 4 : Calculated Isotopic Values Sample Unit 87 Rb/ 86 Sr ( 87 Sr/ 86 Sr) i 147 Sm/ 144 Nd ( 143 Nd/ 144 Nd) i i 176 Lu/ 177 Hf ( 176 Hf/ 177 Hf) i i TOR00 00QR Cuesta Upper Flow 0.8808 0.699665 0.1186 0.511389 3.41 0.0146 0.282168 2.83 TOR00 00QW Channel Lake Member 0.8507 0.707611 0.1335 0.511197 - 0.35 0.0116 0.282212 4.41 TOR00 00RJ Channel Lake Member 0.9977 0.70 5066 0.1362 0.511361 2.86 0.0124 0.282176 3.14 TOR00 00RP Davieaux Island Member 16.8382 0.657469 0.1453 0.511338 2.40 0.0230 0.282098 0.36 TOR00 00RQ South Shore Member 0.5964 0.703865 0.1542 0.511559 6.72 0.0188 0.282186 3.49 TOR00 00S1 South Shore Membe r 0.0423 0.728624 0.1555 0.511016 - 3.89 0.0207 0.281744 - 12.1 8 TOR00 00SD Cuesta Lower Flow 1.7263 0.700482 0.1652 0.511310 1.86 0.0198 0.282048 - 1.42 TOR00 00SJ Quebec Harbor Member unaltered 1.1243 0.703078 0.1270 0.511227 0.23 0.0114 0.282179 3.23 Tabl e 4 ( d) Sample 238 U/ 204 Pb 232 Th/ 204 Pb 235 U/ 204 Pb ( 206 Pb/ 204 Pb) i ( 207 Pb/ 204 Pb) i ( 208 Pb/ 204 Pb) i TOR0000QR 35.48 123.98 0.26 14.0693 15.3243 34.1474 TOR0000QW 16.00 61.69 0.12 17.0266 15.5128 36.8679 TOR0000RJ 15.85 64.33 0.11 16.3006 15.4585 35.8142 TOR0000RP 18.21 170.96 0.13 17.5920 15.5035 33.0358 TOR0000RQ 15.79 64.90 0.11 16.5554 15.4815 35.8315 TOR0000S1 29.81 118.65 0.22 17.3393 15.5059 36.9649 TOR0000SD 13.31 54.95 0.10 16.3127 15.4665 35.5657 TOR0000SJ 19.68 76.03 0.14 19.5291 15.6698 39 .7000 102 APPENDIX C STANDARDS 103 Table 5: Major E lement S tandards Sample SiO2 (%) TiO2 (%) Al2O3 (%) Fe2O3 (%) MnO (%) MgO (%) CaO (%) Na2O (%) K2O (%) P2O5 (%) LOI (%) Sum (%) std JB1a 52.65 1.27 14.46 8.98 0.15 7.82 9.34 2.78 1.43 0.26 0.66 99.14 STD JB 1A 52.51 1.28 14.47 8.99 0.15 7.81 9.34 2.77 1.43 0.26 0.81 99.01 STD JB1A 52.46 1.27 14.48 9.02 0.15 7.83 9.34 2.77 1.43 0.26 0.79 99.01 Jb - 1a GIVEN 52.41 1.28 14.45 9.02 0.148 7.83 9.31 2.73 1.4 0.26 - 98.838 std BHVO - 1 49.59 2.73 13.64 12.18 0.17 7.1 8 11.33 2.2 0.52 0.27 0.05 99.81 STD BHVO - 1 49.61 2.73 13.64 12.21 0.17 7.19 11.34 2.2 0.52 0.27 0 99.88 STD BHVO - 1 49.53 2.73 13.62 12.24 0.17 7.17 11.33 2.19 0.52 0.27 0.1 99.77 BHVO - 1 GIVEN 49.94 2.71 13.8 12.23 0.168 7.23 11.4 2.26 0.52 0.273 - 100. 531 STD RGM - 2 73.77 0.27 13.84 1.86 0.04 0.28 1.16 4.07 4.33 0.05 0.2 99.67 STD RGM - 2 73.74 0.27 13.81 1.86 0.04 0.28 1.16 4.09 4.32 0.05 0.25 99.62 STD RGM - 2 73.55 0.27 13.81 1.87 0.04 0.28 1.16 4.07 4.32 0.05 0.45 99.42 RGM - 2 GIVEN 73.45 0.27 13.72 1 .86 0.036 0.27 1.15 4.07 4.3 0.048 - 99.174 Major elements concentrations were analyzed at Michigan State University on a Bruker S4 Pioneer X - Ray Fluorescence Spectrometer (XRF). 104 Table 6: Trace E lement S tandards Trace elements analyses using a Photon - Machines Analyte G2 Excimer laser and Thermo Scientific ICAP Q quadru pole inductively coupled plasma mass spectrometer (ICP - MS). 105 Table 106 Table 107 Table 108 Table 6 109 110 111 112 113 REFERENCES 114 REFERENCES Annells, R.N., 1974, Keweenawan Volcanic Rocks of Michipicoten Island, Lake Superior, Ontario: An Eruptive Centre of Proterozoic Age, 41N: Department of Energy, Mines and Resources. Annells, 1973, Proterozic Flood Basalt of eastern Lake Super i or: The Keweenawan Volcanic Rocks of the Mamainse Point Area, Ontario: Department of Energy, Mines and Resource s Calgary, Alberta, Canada, v. 5. Bachmann, O., and Bergantz, G.W., 2004, On the origin of crystal - poor rhyolites: extracted from batholithic cr y stal mushes: Journal of Petrology, v. 45, p. 1565 1582. Baragar, W.R.A., 1978, Michipicoten Island, Ontario: Ru bidium - strontium isotopic age studies, Report, v. 2, p. 77 14. Beard, J.S., and Lofgren, G.E., 1989, Effect of water on the composition of parti a l melts of greenstone and amphibolite: Science, v. 244, p. 195 197. Bell, K., and Blenkinsop, J., 1987, Archean depleted mantle: evidence from Nd and Sr initial isotopic ratios of carbonatites: Geochimica et Cosmochimica Acta, v. 51, p. 291 298. Bennett, V .C., 2003, Compositional evolution of the mantle: Treatise on geochemistry, v. 2, p. 568. Blichert - Toft, J., Ch auvel, C., and Albarède, F., 1997, Separation of Hf and Lu for high - precision isotope analysis of rock samples by magnetic sector - multiple colle c tor ICP - MS: Contributions to Mineralogy and Petrology, v. 127, p. 248 260. Bouvier, A., Vervoort, J.D., and Pat chett, P.J., 2008, The Lu Hf and Sm Nd isotopic composition of CHUR: constraints from unequilibrated chondrites and implications for the bulk co m position of terrestrial planets: Earth and Planetary Science Letters, v. 273, p. 48 57. Bryan, S.E., and Ernst, R.E., 2008, Revised definition of large igneous provinces (LIPs): Earth - Science Reviews, v. 86, p. 175 202. Buck, W.R., and Karner, G.D., 2004, Consequences of asthenospheric variability on continental rifting: Rheology and deformation of the lithosphere at continental margins, v. 62, p. 1 30. Cannon, W.F., 1994, Closing of the Midcontinent rift - A far field effect of Grenvillian compression: Geol o gy, v. 22, p. 155 158. Cannon, W.F., 1992, The Midcontinent rift in the Lake Superior region with emphasis on i ts geodynamic evolution: Tectonophysics, v. 213, p. 41 48. Cannon, W.F., and Hinze, W.J., 1992, Speculations on the origin of the North American Midcontinent rift: Tectonophysics, v. 213, p. 49 55. 115 Chaudhuri, S., and Faure, G., 1967, Geochronology of the K eweenawan rocks, White Pine, Michigan: Economic Geology, v. 62, p. 1011 1033. Chauvel, C., Lewin, E., Carpentier, M., Arndt, N.T., and Marini, J . - C., 2008, Role of recycled oceanic basalt and sediment in generating the Hf Nd mantle array: Nature geoscience , v. 1, p. 64. Chiarenzelli, J.R., and McLelland, J.M., 1991, Age and regional relationships of granitoid rocks of the Adirondack Highlands: The Journal of Geology, v. 99, p. 571 590. Connelly, J.N., Ulfbeck, D.G., Thrane, K., Bizzarro, M., and Housh, T., 2006, A method for purifying Lu and Hf for analyses by MC - ICP - MS using TODGA resin: Chemical Geology, v. 233, p. 126 136. Courtillot, V., Jaupar t , C., Manighetti, I., Tapponnier, P., and Besse, J., 1999, On causal links between flood basalts and continental breakup: Earth and Planetary Science Letters, v. 166, p. 177 195. Davis, D.W., and Green, J.C., 1997, rift in western Lake Superior and implic a tions for: Earth, v. 488, p. 476 488. Davis, D.W., and Paces, J.B., 1990, Time resolution of geologic events on the Keweenaw Peninsula and implications for development of the Midcontinent Rift system: Earth and Planetary Science Letters, v. 97, p. 54 64. D eering, C.D., and Bachmann, O., 2010, Trace element indicators of crystal accumulation in silicic ign eous rocks: Earth and Planetary Science Letters, v. 297, p. 324 331. Dixon, S., and Rutherford, M.J., 1979, Plagiogranites as late - stage immiscible liquid s in ophiolite and mid - ocean ridge suites: an experimental study: Earth and Planetary Science Letters, v. 45, p. 45 60. Dosso, L., 1984, The nature of the Precambrian subcontinental mantle: Isotopic study (Sr, Pb, Nd) of the Keweenawan volcanism of the Nor t h Shore of Lake Superior: University of Minnesota. Ernst, R.E., 2014, Large igneous provinces: Cambri dge University Press. Fairchild, L.M., Swanson - Hysell, N.L., Ramezani, J., Sprain, C.J., and Bowring, S.A., 2017, The end of Midcontinent Rift magmatism a n d the paleogeography of Laurentia: Lithosphere, v. 9, p. 117 133. Faure, G., 1986, Principles of isot ope geology, 2nd edn: Wiley. New York. Finlayson, V.A., Konter, J.G., Konrad, K., Koppers, A.A.P., Jackson, M.G., and Rooney, T.O., 2018, Sr Pb Nd Hf isot o pes and 40Ar/39Ar ages reveal a Hawaii Emperor - style bend in the Rurutu hotspot: Earth and Planetary Science Letters, v. 500, p. 168 179. Good, D., and Lightfoot, P.C., 2019, Significance of Metasomatized Lithospheric Mantle in the Formation of Early Basa l ts and Cu - PGE Sulfide Mineralization in the Coldwell Complex, Midcontinent Rift, Canada: Canadian Jou rnal of Earth Sciences,. 116 Grant, J.A., 2005, Isocon analysis: A brief review of the method and applications: Physics and Chemistry of the Earth, Parts A/B/ C , v. 30, p. 997 1004. Gualda, G.A., Ghiorso, M.S., Lemons, R.V., and Carley, T.L., 2012, Rhyolite - MEL TS: a modified calibration of MELTS optimized for silica - rich, fluid - bearing magmatic systems: Journal of Petrology, v. 53, p. 875 890. Heaman, L.M., East o n, R.M., Hart, T.R., MacDonald, C.A., Hollings, P., and Smyk, M., 2007, Further refinement to the tim ing of Mesoproterozoic magmatism, Lake Nipigon region, Ontario: Canadian Journal of Earth Sciences, v. 44, p. 1055 1086. Helz, R.T., 1976, Phase relations of basalts in their melting ranges at P H2O= 5 kb. Part II. Melt compositions: Journal of Petrology, v. 17, p. 139 193. Hess, P.C., 1971, Polymer model of silicate melts: Geochimica et Cosmochimica Acta, v. 35, p. 289 306. Hildreth, W., 2004, Volcanologic a l perspectives on Long Valley, Mammoth Mountain, and Mono Craters: several contiguous but discrete sy stems: Journal of Volcanology and Geothermal Research, v. 136, p. 169 198. Hollings, P., Hart, T., Richardson, A., and MacDonald, C.A., 2007, Geochemistry of the Mesoproterozoic intrusive rocks of the Nipigon Embayment, northwestern Ontario; evaluating the earliest phases of rift development: Canadian Journal of Earth Sciences = Revue Canadienne des Sciences de la Terre, v. 44, p. 1087 1110, doi:http://dx.d o i.org/10.1139/E06 - 127. Irvine, T.N.J., and Baragar, W.R.A., 1971, A guide to the chemical classificat ion of the common volcanic rocks: Canadian journal of earth sciences, v. 8, p. 523 548. - F., Martin, H., Erban, V., and Farrow, C., 2016, Geochemical Modelling of Igneous Processes Principles And Recipes in R Language: Bringing the Power of R to a Geochemical Community: Berlin Heidelberg, Springer - Verlag, Springer Geochemistry, https://www.springer.com/us/book/9783662467916 (accesse d April 2019). Klewin, K.W., and Berg, J.H., 1990, Geochemistry of the Mamainse Point volcanics, Ontar io, and implications for the Keweenawan paleomagnetic record: Canadian Journal of Earth Sciences, v. 27, p. 1194 1199. Klewin, K.W., and Berg, J.H., 1991, Petrology of the Keweenawan Mamainse Point lavas, Ontario: Petrogenesis and continental rift evolutio n: Journal of Geophysical Research: Solid Earth, v. 96, p. 457 474. Konter, J.G., and Storm, L.P., 2014, High precision 87Sr/86Sr measurements by MC - ICP - M S , simultaneously solving for Kr interferences and mass - based fractionation: Chemical Geology, v. 385, p. 26 34, doi:10.1016/j.chemgeo.2014.07.009. Krans, S.R., Rooney, T.O., Kappelman, J., Yirgu, G., and Ayalew, D., 2018, From initiation to termination: a petrostratigraphic tour of the Ethiopian Low - Ti Flood Basalt Province: Contributions to Mineralogy an d Petrology, v. 173, p. 37. 117 de Laeter, J.R., Böhlke, J.K., De Bièvre, P., Hidaka, H., Peiser, H.S., Rosman, K.J., and Taylor, P.D., 2003, Atomic weights o f the elements. Review 2000 (IUPAC Technical Report): Pure and applied chemistry, v. 75, p. 683 800. L e Bas, M.J., Le Maitre, R.W., Streckeisen, A., Zanettin, B., and Rocks, I.S. on the S. of I., 1986, A chemical classification of volcanic rocks based on t h e total alkali - silica diagram: Journal of petrology, v. 27, p. 745 750. Lee, C. - T.A., Lee, T.C., and Wu, C. - T., 2014, Modeling the compositional evolution of recharging, evacuating, and fractionating (REFC) magma chambers: Implications for differentiation of arc magmas: Geochimica et Cosmochimica Acta, v. 143, p. 8 22. Marshall, L., 1996, Geochemistry and paleomagnetism of Keweenawan basalt in the subsurface of Nebraska: Precambrian Research, v. 76, p. 47 65, doi:10.1016/0301 - 9268(95)00027 - 5. Matthews, T.P . , and Rooney, T.O., 2009, Deducing Rhyolite Origins in Keweenawan Magmas - Geochemical Correlations Wit h Flood Basalts, Mamainse Point, ON, in AGU Fall Meeting Abstracts,. McBirney, A.R., 1996, The Skaergaard intrusion, in Developments in Petrology, Elsevie r , v. 15, p. 147 180. McDonough, W.F., and Sun, S. - S., 1995, The composition of the Earth: Chemical geology, v. 120, p. 223 253. McLelland, J., Hamilton, M., Selleck, B., McLelland, J., Walker, D., and Orrell, S., 2001, Zircon U - Pb geochronology of the Ott a wan orogeny, Adirondack highlands, New York: regional and tectonic implications: Precambri an Research, v. 109, p. 39 72. Miller, C.F., and Mittlefehldt, D.W., 1982, Depletion of light rare - earth elements in felsic magmas: Geology, v. 10, p. 129 133. Mille r , J., and Nicholson, S.W., 2013, Geology and mineral deposits of the 1.1 Ga Midcontinent R ift in the Lake Superior region an overview: Field guide to the copper - nickel - platinum group element deposits of the Lake Superior Region. Edited by Miller, J. Preca m brian Research Center Guidebook, v. 13, p. 1 49. Nicholson, S.W., 1990, Portage Lake rhyol ites of the midcontinent rift system, Keweenaw Peninsula, Michigan: geology, petrogenesis and implications for rift magmatism and mineralization: University of Minne s ota Minneapolis. Nicholson, S.W., Schulz, K.J., Shirey, S.B., and Green, J.C., 1997, Rift - wide correlation of 1.1 Ga Midcontinent rift system basalts: implications for multiple mantle sources during rift development: Canadian Journal of Earth Sciences, v. 34, p. 504 520. Nicholson, S.W., and Shirey, S.B., 1990, Midcontinent rift volcanism in th e Lake Superior region: Sr, Nd, and Pb isotopic evidence for a mantle plume origin: Journal of Geophysical Research: Solid Earth, v. 95, p. 10851 10868. Paces, J.B., 1988, Magmatic processes , evolution and mantle source characteristics ...: 118 Paces, J.B., a nd Bell, K., 1989a, Non - depleted sub - continental mantle beneath the Superior Province of the Canadian Shield: Nd - Sr isotopic and trace element evidence from Midconti n ent Rift basalts: Geochimica et Cosmochimica Acta, v. 53, p. 2023 2035. Paces, J.B., and B ell, K., 1989b, Non - depleted sub - continental mantle beneath the Superior Province of the Canadian Shield: Nd - Sr isotopic and trace element evidence from Midcontinent Rift basalts: Geochimica et Cosmochimica Acta, v. 53, p. 2023 2035, doi:10.1016/0016 - 7037( 89)90322 - 0. Palmer, H.C., and Davis, D.W., 1987, Paleomagnetism and U - Pb geochronology of volcanic rocks from michipicoten island, lake superior, canada: precise cal i bration of the keweenawan polar wander track: Precambrian Research, v. 37, p. 157 171, doi :10.1016/0301 - 9268(87)90077 - 5. Philpotts, A.R., 1982, Compositions of immiscible liquids in volcanic rocks: Contributions to Mineralogy and Petrology, v. 80, p. 201 2 18. Polat, A., and Kerrich, R., 2000, Archean greenstone belt magmatism and the continenta l growth mantle evolution connection: constraints from Th U Nb LREE systematics of the 2.7 Ga Wawa subprovince, Superior Province, Canada: Earth and Planetary Scienc e Letters, v. 175, p. 41 54. Polat, A., and Münker, C., 2004, Hf Nd isotope evidence for co ntemporaneous subduction processes in the source of late Archean arc lavas from the Superior Province, Canada: Chemical Geology, v. 213, p. 403 429. Rivers, T., 2008 , Assembly and preservation of lower, mid, and upper orogenic crust in the Grenville Provin ce Implications for the evolution of large hot long - duration orogens: Precambrian Research, v. 167, p. 237 259. Rivers, T., 1997, Lithotectonic elements of the Grenv i lle Province: review and tectonic implications: Precambrian Research, v. 86, p. 117 154, d oi:10.1016/S0301 - 9268(97)00038 - 7. Rivers, T., Ketchum, J., Indares, A., and Hynes, A., 2002, The High Pressure belt in the Grenville Province: architecture, timing, a nd exhumation: Canadian Journal of Earth Sciences, v. 39, p. 867 893. Rooney, T.O., LaVign e, A., Konter, J., Eric, B., Stein, S., Moucha, R., and Stein, C., 2018, CONSTRAINING THE GEOCHEMICAL RESERVOIRS CONTRIBUTING TO VOLCANISM IN THE KEWEENAW LIP, in GS A , https://gsa.confex.com/gsa/2018AM/webprogram/Paper323279.html (accessed June 2019). Saun ders, A.D., England, R.W., Reichow, M.K., and White, R.V., 2005, A mantle plume origin for the Siberian traps: uplift and extension in the West Siberian Basin, Russi a : Lithos, v. 79, p. 407 424. Shirey, S.B., 1997, Re - Os isotopic compositions of Midcontine nt rift system picrites: implications for plume lithosphere interaction and enriched mantle sources: Canadian Journal of Earth Sciences, v. 34, p. 489 503. 119 Shirey, S . B., Berg, J.H., and Carlson, R.W., 1994, Temporal changes in the sources of flood basalts: isotopic and trace element evidence from the 1100 Ma old Keweenawan Mamainse Point Formation, Ontario, Canada: Geochimica et Cosmochimica Acta, v. 58, p. 4475 4490. Spulber, S.D., and Rutherford, M.J., 1983, The origin of rhyolite and plagiogranite in oce anic crust: an experimental study: Journal of Petrology, v. 24, p. 1 25. t inent Rift: when rift met LIP: Geosphere, v. 11, p. 1607 1616. Stein, S., Stein, C.A., Ell ing, R., Kley, J., Keller, R., Wysession, M., Rooney, T., Frederiksen, A., evol u tion of continental rifts and passive continental margins: Tectonophysics,. Stein, C.A., S tein, S., Merino, M., Randy Keller, G., Flesch, L.M., and Jurdy, D.M., 2014, Was the Midcontinent Rift part of a successful seafloor spreading episode? Geophysical R esearch Letters, v. 41, p. 1465 1470. Swanson - Hysell, N.L., Burgess, S.D., Maloof, A.C., an d Bowring, S.A., 2014, Magmatic activity and plate motion during the latent stage of Midcontinent Rift development: Geology, v. 42, p. 475 478. Swanson - Hysell, N.L., Ramezani, J., Fairchild, L.M., and Rose, I.R., 2019, Failed rifting and fast drifting: Mid Grenvillian orogenesis: Geological Society of America Bulleti n,. Thy, P., Beard, J.S., and Lofgren, G.E., 1990, Experimental Constraints on the Origin of Icelandic Rhyolites: The Journa l of Geology, v. 98, p. 417 421. Turek, A., Smith, P.E., and Schmus, W.V., 1982, Rb Sr and U Pb ages of volcanism and granite emplac ement in the Michipicoten belt Wawa, Ontario: Canadian Journal of Earth Sciences, v. 19, p. 1608 1626. Vervoort, J.D., and G reen, J.C., 1997, Origin of evolved magmas in the Midcontinent rift system, northeast Minnesota: Nd - isotope evidence for melting of Archean crust: Canadian Journal of Earth Sciences, v. 34, p. 521 535. Vervoort, J.D., Wirth, K., Kennedy, B., Sandland, T., and Harpp, K.S., 2007, The magmatic evolution of the Midcontinent rift: New geochronologic and geochemical evidence from felsic magm atism: Precambrian Research, v. 157, p. 235 268. White, R., and McKenzie, D., 1989, Magmatism at rift zones: the genera tion of volcanic continental margins and flood basalts: Journal of Geophysical Research: Solid Earth, v. 94, p. 7685 7729. Wirth, K.R., N aiman, Z.J., and Vervoort, J.D., 1997, The Chengwatana Volcanics, Wisconsin and Minnesota: petrogenesis of the southern most volcanic rocks exposed in the Midcontinent rift: Canadian Journal of Earth Sciences, v. 34, p. 536 548. 120 Workman, R.K., and Hart, S.R., 2005, Major and trace element composition of the depleted MORB mantle (DMM): Earth and Planetary Science Letters, v. 231, p. 53 72. Zindler, A., and Hart, S., 1986, Chemical Geodynamics: Annual Review of Earth and Planetary S ciences, v. 14, p. 493 571, doi:10.1146/annurev.ea.14.050186.002425.