..... ,,~.~ ", ' .. /.~ c ':t-, ).;' J THE PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY OF MINNESOTA UNIVERSITY OF MINNESOTA MINNESOTA GEOLOGICAL SURVEY GEORGE M. SCHWARTZ, DIRECTOH BULLETIN 41 The Precambrian Geology and Geochronology of Minnesota BY SAMUEL S. GOLDICH, ALFRED O. NIER, HALFDAN BAADSGAARD, JOHN H. HOFFl\lAN, AND HAROLD W. KRUEGER MINNEAPOLIS • 1961 THE UNIVERSITY OF MINNESOTA PRESS PRINTED IN THE UNITED STATES OF AMERICA AT THE LUND PRESS, INC., MINNEAPOLIS "~O Library of Congnss Catalog Card Number: 61-8016 PUBLISHED IN GREAT BRITAIN, INDIA, AND PAKISTAN BY THE OXFORD UNIVERSITY PRESS LONDON, BOl\lIBAY, AND KARACHI, AND IN CANADA BY THOMAS ALLEN, LTD., TORONTO FOR.EWOR.D This bulletin is an outstanding example of the cooperation of several scientists and scientific organizations. Without the cooperation of all con- cerned, such a comprehensive correlation of age determinations and re- gional geology would have been impossible short of many years of work. The results are a particularly important demonstration of cooperation among geologists, chemists, and physicists. The grants from the National Science Foundation were a major source of funds and made the project possible. In addition the Minnesota Geolog- ical Survey supported the project, particularly the field work and the entire cost of publication. The Graduate School of the University of Min- nesota also made notable contributions of funds at critical points during the course of the project. As the authors point out in several places, geochronology is a relatively new development and much is still to be learned. The present work is, in effect, a pioneering effort and points the way to more work in JVlinnesota and elsewhere. As Director of the Minnesota Geological Survey, I wish to express my sincere appreciation of their devoted work to all concerned, especially to Dr. Samuel S. Goldich who initiated the project and furnished the extra energy to push it to a conclusion. George M. Schwartz v ACKNOWLEDGMENTS The research on which this report is based was made possible by grants from the National Science Foundation and from the Graduate School of the University of Minnesota. We are also grateful to the Minnesota Geo- logical Survey for its support and to G. M. Schwartz, Director, for his active cooperation and interest. The assistance of our colleagues in the Department of Geology and Mineralogy and in the School of Physics con- tributed materially to the progress of the investigation. The School of Mines and Metallurgy and the Department of Chemical Engineering gen- erously permitted use of their crushing and grinding equipment which greatly facilitated the preparation of the samples. We are indebted to D. H. Yardley of the School of Mines and Metallurgy for his contributions to the work in Ontario. The field notes of F. F. Grout, J. W. Gruner, and others, who in past years were active in the work of the Survey, were a valuable source of information as were the field projects of former graduate students of the Department of Geology and Mineral- ogy. Numerous samples were contributed by geologists of mining com- panies, and these are acknowledged in the text and in the Appendix. The interest shown by the geologists of the Lake Superior region in theprogress of the investigation and their helpful discussions of many problems are much appreciated. P. W. Gast of the Department of Geology and Mineralogy gave valuable assistance in the rubidium-strontium dating. Much of the chemical work was done by Daniel H. Anderson, and most of the rubidium and strontium isotope measurements were performed by Robert Stryk. Donald Maund made many of the argon isotope measurements, and we are indebted to Doris Thaemlitz, Eileen H. Oslund, and C. O. Ingamells for the potassium determinations which were made in the Rock Analysis Laboratory. Z. E. Peterman, .J. K Frye, David Alt, Curtis Bury, and Donald Keairnes gave valuable assistance in the field and in the laboratory. Gil- bert Hanson, Lawrence Mangen, and Marshall Swain assisted with the sample preparation. X-ray studies were made by Anthony Tennissen. R. L. Buchheit, P. M. Pilling, and John Zimmer did most of the drafting. Special thanks are due R. B. Thorness of the Physics Research Shop who supervised and planned the construction of the mass spectrometers and related apparatus. The final revision of the manuscript was completed in Washington, D.C., and we are grateful for suggestions made by members of the United States Geological Survey. l\tlany persons have read and offered comments on parts or all of the Vll viii PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY manuscript; however, the authors are responsible for the views expressed and for any shortcomings not only in the experimental data and inter- pretations but in their presentation as well. It is a pleasure to acknowledge the helpful suggestions and criticisms of Wenonah Helen Bergquist, R. A. Burwash, P. E. Cloud, Jr., C. E. Dutton, R. E. Folinsbee, Henry Lepp, Joseph Lipson, R. W. Marsden, G. M. Schwartz, and D. H. Yardley. CONTENTS FOREWORD.................................................... V ACKNOWLEDGMENTS ........................................... VII ABSTRACT ..................................................... XIX 1. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . .. . ................... . The Problem ............................................. . Presentation of Results ................................. . ~. METHODS ............................................. . General Statement ... . Potassium-Argon Method 1 fl 3 8 8 8 Development ................................. 8 Sample Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Determination of Potassium ... . . . . . . . . . . . . . . . . . . . . . 10 J. Lawrence Smith Method . . . . . . . . . . . . . . . . . . . . . . 10 Blank. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Rapid Flame Method .. 11 Precision and Accuracy . . . . . . . . . . . . . . . . . . 11 Determination of Argon . . . . . . . . . . . . . . . . . . . . . . . . 13 Extraction .............................. 13 A'lB Spikes ...................... 15 Isotopic Measurements .............................. . 16 Analytical Errors ............. . Ru bidium-Strontium Method Development ........... . . . . . . . . . . . . . . . . . . . . . .. 21 21 21 Sample Preparation . . . . . . . . . . . . fl2 Analytical Procedures . . . . . . . . . . . . . flfl Reagents and Apparatus. . . . . . . . . . . . . . . . . . . . fl2 Decomposition .................................... flfl Rubidium ............................... fl2 Strontium ................................ fl3 Spike Solutions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 IX x PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY Isotopic :Measurements ... . Analytical Errors ........... . Evaluation of Radioactivity Ages. Analytical Problems .......... . Bikita Quarry Lepidolite ...... . Chestnut Flat Mine Muscovite .... . Petrick Quarry Biotites .......... . Comparison of K-A and Rb-Sr Ages .. Decay Constants ..................... . Geologic Problems ........................ . 8. NORTHERN MINNESOTA Introductory Statement ............ . Vermilion District .................. . General Statement ........................ . Keewatin Group ......... . Ely Greenstone ......... . Soudan Iron-Formation Laurentian Intrusives .... . Dikes .................. . Northern Light Gneiss .. Saganaga Granite .............. . Younger Intrusives ..... Sequence of Intrusions . Knife Lake Group ...... . History of the Name .. . Ogishke Conglomerate Agawa Iron-Formation. Principal Rock Types .. . Structure and Thickness .... . Ages for the Vermilion District Soudan Area ................ . Knife Lake Area ..... . Snowbank Lake Area .. Saganaga Lake Area ... Vermilion Granite Region .. General Statement ... . Vermilion Granite ...... . .' ..... , .. 23 24 25 25 25 30 31 36 86 40 40 41 41 4'2 44 44 44 46 46 48 49 ·1-9 49 49 50 50 51 51 51 54 57 .57 58 CONTENTS xi Burntside Granite Gneiss .. . ....................... " 58 Other Gneissic Rocks .... . Pegmatites ..................... . Ages for the Vermilion Granite Region Vermilion Granite ................ . Northwest Angle Pegmatites ......... . Giants Range Region ...... . General Statement ....... . 59 60 60 60 62 . ...... 62 62 Giants Range Granite . . . . . . . . . . . . ()3 Accessory-Mineral Zonation ..... 64 Relative Ages from Zircon Types 64 Ages for the Giants Range Region. . . . 65 History of Early Precambrian Time for Northern Minnesota 65 General Statement ............................ . Geologic Interpretation ... Coutchiching Problem .. Keewatin Group ..... . Laurentian Orogeny ............ . Knife Lake Group ............ . Algoman Orogeny ............. . Geochronology ............................... . Data from Ontario ...................... . Analysis of the Data ............ . Summary Statement ....................... . Mesabi District .............................. . General Statement ........... . Animikie Group ....................... . Type Locality. . . . . . . . ............ . Pokegama Formation ............... . Biwabik Formation ....... . Virginia Formation .. . . . . . . .. ()5 6() ()6 67 67 68 ()8 69 ()9 71 73 74 74 75 75 75 76 Relationships ...................................... . 76 77 78 78 Intrusive Rocks ..................................... . Age of the Animikie Group .............................. . History of Middle Precambrian Time for Northern Minnesota. .. 78 Eparchean Interval .................................... " 78 North Shore Region .............................. ' ........ " 79 xu PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY General Statement ........................ . Thomson Formation ...................... . Puckwunge Formation ...... . North Shore Volcanic Group . . . . . . ....... . Duluth Complex .......................... . Early Work ............................ . Multiple Intrusions ..................... . Ferrogranodiorite and Granophyre ............ . Dikes ............................... . Origin of the Gabbro Complex at Duluth. Duluth Complex in Northern Areas ..... . Sills at Duluth ............................... . Endion Sill .......................... . Northland Sill ............................... . Lester River Sill .... Beaver Bay Complex .. Logan Intrusives ......................... . 79 79 79 81 82 82 8~) 84 85 86 89 90 90 90 91 91 91 Origin of the Keweenawan Igneous Rock Series of Minnesota .. 92 Upper Keweenawan Sandstones. . . . . . . . . .......... 9$ Fond du Lac Formation . . . . . . . . . . . . . . . 9:) Hinckley Formation ................ 9-1 Ages for the North Shore Region . . . . . . . . . . . . . 95 Duluth Complex and Related Intrusives. . . . 95 Contact Rocks .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 Granophyre in the Duluth Area. . . . . . . . . . . . . . . . . 9G History of Late Precambrian Time for Northern Minnesota. . .. 96 Lower Keweenawan. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9G Middle Keweenawan . . . . . . . . . . . . . . . . . . . . . . . 97 Lake Superior Geosyncline ... . Duluth Complex ............ . Upper Keweenawan ....... . 97 98 99 4. EAST-CENTRAL lVIINNESOTA .... · .101 Introductory Statement . Thomson Formation ... General Description .............. . Description of Units .......... . Gra~'wacke and Slate at Type Locality .. · .101 . ...... 102 · .102 . ..... 105 · .105 CONTENTS XliI Phyllite at Atkinson ................................... 106 Phyllite at Barnum . . . . . . . . . . . . . . . . . . . . ...... 106 Schist at Moose Lake. . . . . . . . . . . . . . . . . . .. 106 Paragneiss and Schist at Denham. . . . . . . . .106 Phyllite at Little Falls .................................. 107 Ages for Thomson Metasediments . . . . . . . . . . . . . . ... 108 McGrath Gneiss. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 108 Description ..................................... . ... 108 Age. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . .. 109 Metasediments of the Cuyuna District .. 110 Description ................................. . .110 Samples and Age Determinations .......................... 110 Igneous Rocks ............................................ III General Description ...................................... III Older Granodiorites and Related Rocks. . . . . . . . . . . . . . . .. 111 St. Cloud Gray Granodiorite . . . . . . . . . . ... III Warman Quartz Monzonite .. Hillman Tonalite ......... . Freedhem Tonalite ........ . Isle Quartz Monzonite ...... . Younger Granites ........ . St. Cloud Red Granite ........ . Rockville Porphyritic Granite ...... . Penokean Orogeny ................. _ .. . General Statement ........... _ Thomson Formation Problem .. . The Granite Problem ..... . Middle Precambrian History .... . Summary for Minnesota ...... . Correlation with Other Districts .... _ Wisconsin ................. _ Michigan .................. . .5. SOUTHWESTERN l\![INNESOTA ....... . Introductory Statement .................. . Morton-Sacred Heart Region .... . General Statement ........... . . ....... 112 . ... 112 · .112 · .112 · .113 . ....... 113 _ .. 113 _ .. 114 _ .. 114 · .114 - ... 116 · .117 . ... 117 ... 120 . _ .121 · .121 . .... 123 . ...... 123 · .12.5 . - ......... 12.5 XIV PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY Basic Complex ........ . ............................... 127 :Morton Quartz Monzonite Gneiss Seaforth Gneiss ..... Sacred Heart Granite ........ . Granite of Larsen Quarry .. . Fort Ridgley Granite ......... . Post-Morton Gneiss Intrusives .. Cedar Mountain Complex Granophyre Gabbro ..... . ....... . Granite ...................... . Age and Origin ................... . Granite Falls-Montevideo Region ....... . General Statement .................. . Garnetiferous Quartz Diorite Gneiss ... . Gabbro and Diorite Gneiss .......... . Montevideo Granite Gneiss ....... . Description ............... . Composition .................. . Structure .................. . Basalt Dikes ................ . Late Granite of Section 28 .... . Age and Origin ............ . · .. 128 · .... 129 · .. 129 · .... 130 . ........ 131 · ... 131 ................ 131 . ..... 133 · .... 133 · .134 . ..... 137 · .137 · .138 · .... 138 · .. 139 · .... 139 · .. 1~39 · ... 139 · .... 140 · ... 140 · .... 142 Odessa-Ortonville Region ................................... 144 General Statement. . . . . . . . . . . . ... 144 Basic Complex . . . . . . . . . . . . . . .. 144 Odessa Granite. . . .. ........ . ..... 145 Bellingham Granite .......... . · ... 146 Milbank Granite .. · ... 146 Age and Origin · .... 146 Sioux Formation .......................................... 147 Description .......... . .. 147 Pipestone or Catlinite . . . . . . . . . . . . .... 14!:l Age and Correlation. . . . ... 149 6. DEVELOPMENT OF A PRECAIVIBRIAN CLASSIFICATION · .... 150 Introductory Statement ....... . · ... 150 Time Scale .............. . . ...... 151 CONTENTS xv Principles .............................................. 1.51 Orogenies ........ . Unconformities .... . Correlation ....... . Nomenclature .............................. . Early Precambrian ................... . ............ 151 · .... 152 · .... 152 . ....... 153 . ........ 153 Ontarian ................................. . . ....... 153 Laurentian Orogeny .................... . ... 154 Timiskamian ...... . · ... 154 Algoman Orogeny .. . . ...... 154 Correlation ....................................... . .. 155 Middle Precambrian . . . . . . . . . . . . . . . . . . . . . . . . .. . .. 155 Huronian ............................... . · ... 156 Penokean Orogeny .............................. . · ... 156 Correlation ............................. . . ..... 1.57 Older Cycles of the ·Middle Precambrian .... . · .... 159 Late Precambrian ....................................... 160 Keweenawan .......................................... 160 Grenville Orogeny ..... . . . . . . . . . . . . . . ..... 161 Correlation ......................... . ..... 161 1.2-1.6 b.y. Ages. . . . . . . . . . . . . . . . . . . . . . ..... 162 Younger Cycles ......................... . ... 163 Killarney Revolution. . . . . . . . . . . . . . . . . . . . . .. 163 Cambrian-Precambrian Time Boundary · .... 164 Concluding Remarks .............. . · ... 165 ApPENDIX .............. , .. ' ... . · .... 171 RE~'ERENCES .......... . . .. 181 INDEX ....... . . ........... 189 LIST OF FIGURES 1. The Canadian Shield, showing position of the State of Minnesota 2 2A. Early argon extraction system. . . . . . . . . . . . . . . . . . . . . . . . 14 2B. Later, modified argon extraction system 14 ~~. Manifold for calibrating A 38 spikes .... 16 4. Schematic diagram of argon mass spectrometer tube .......... . 17 XVI PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY 5. Photograph of mass spectrometer tube. . . . . . . . . . . . . . . . . 18 6. Strip chart of argon from sample KA-24. . . . . . . . . . . . . 19 7. Plot of K-A versus Rb-Sr ages. . . . . . . . . . . . . . . . . . . 27 8. Plot of A4CJ/K40 ratios for mica-feldspar pairs.... . ........... .. 32 !). Plot of K-A versus Rb-Sr ages for micas from rocks of the Appa- lachian Mountains and Sudbury, Ontario, regions. . . . . . . . . . . . 34 J O. Schist and paragneiss, Coutchiching ..................... . 37 1 1. Ely greenstone .......................... 39 1 '2. Tight folds in the Soudan iron-formation. . . . . . . . . 43 IS. Map of the Saganaga batholith. . . . . . . . . . . 45 14. Geologic map of the Snowbank Lake area. . . . 56 15. Crenulated folds and boudin structure. . . . . . . . 60 16. Pebbles and cobbles of Laurentian tonalite. " . .. 61 17. Conglomerate of the Early Precambrian Knife Lake group. . . 63 18. Late Precambrian Puckwunge (?) conglomerate.. 80 19. Rhythmic layering in troctolite. . . . . . . . . . . . . 8':t ':to. Photomicrograph of anorthositic gabbro. . . 84 Q1. Photomicrograph of medium-grained adamellite. . . 85 QQ. Iron-enrichment diagram .... . . .. ........... 88 '23. Alkali-enrichment diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 89 '24. Iron-enrichment diagram for Keweenawan igneous rock series.. 93 '2.5. Schematic diagrams illustrating the development of the Penokean orogeny ......................................... . 26. Index map of the Minnesota River Valley .. . 27. Map of the Morton area ...... ' ........... ' 119 1Q4 1'26 28. Structure in the :Morton gneiss around an inclusion. . . . 1':t7 Q9. Geologic map of the Minnesota River Valley south of Sacred Heart ................................................... 128 30. Geologic map of the Cedar Mountain area. . . . . . . . . . . . . . . . . . 132 31. Igneous agglomerate formed by intrusion of granite into basalt .. 141 32. Close-up view of agglomerate. . . . . . . . . . . . . . . . . . . . . . 141 33. l\Iap of the Minnesota River Valley southeast of Odessa. . . 14.5 CONTENTS xvii ~4. Map showing location of dated samples of pipestone from the Sioux formation and of Cambrian glauconite samples .......... 147 35. Bed of pipestone in quartzite of the Sioux formation. . . . . . . . . .. 148 36. Comparison of radioactivity ages for the Penokean orogenic belt of the Lake Superior region with ages for the Svecofennidic and Karelidic orogenic belts of Finland. . . . . . . . . . . . . . . . . . . . . . . . .. 159 37. Map showing counties of Minnesota ......................... 172 LIST OF PLATES (in pocket) 1. Geological map of northeastern Minnesota '2. Geological map of east-central Minnesota 3. Map of the Minnesota-Ontario border region, showing location of dated samples 4. Map of western part of the Lake Superior region, showing location of dated samples 5. Geological map of the Granite Falls area, showing location of dated samples LIST OF TABLES 1. Precambrian Column in Minnesota, Chronologic and Strati- graphic Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2. Stratigraphic Succession and Geochronology of the Precambrian of Minnesota ............................................. 5 3. Comparison of Results for K 20 by Rapid Flame Spectrophoto- metric and Gravimetric Methods .... " . . . . . . . . . . . . . . . . . . . . .. 12 4. Replicate Determinations for K 20 on Granite G-1. . . . . . . . . . . .. 12 5. Replicate Determinations of K 20, Rb20, and Cs20 on Lepidolites 12 6. Analytical Data for Lepidolite, Muscovite, and Biotite from In- terlaboratory Check Samples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 26 7. Sr87/Rb87 Age Determinations. . . . . . . . . . . . . . . . . . . . . . . . . .. ... 28 8. A 40 /K40 Age Determinations for Miscellaneous Samples. . . . . 29 9. Comparative Ages Obtained by Different Dating Methods. . . . 31 10. Precambrian Succession and Nomenclature Recommended by Special Committee ....................................... '. 36 XVlll PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY 11. ~equenc~ and Classification of the Precambrian of the Lake Supe- rIor RegIOn . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 38 12. Stratigraphic Succession in the Vermilion District. . . . . . . . . . .. 41 13. Chemical Analyses of Saganaga Granite, Northern Light Gneiss, and Gneiss from Rice Bay. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 47 14. Ages for Rocks from Northern Minnesota (Including Data from Ontario) ................................................. 52 15. Chemical Analyses of Ely Greenstone, Wall Rock Alteration, Seri- cite Schist, and Sericite Samples from Soudan Area. . . . . . . . . . . .. 55 16. Chemical Analyses of Vermilion, Giants Range, and Snowbank Granites, and of Gneiss from Burntside Lake. . . . . . . . . . . . . . . . .. 58 17. Chemical Analyses of Ferrogranodiorite, Granophyre, and Re- lated Types from Duluth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 86 18. Chemical Analyses of Basaltic Rocks from the Duluth Area. . . .. 87 19. Geologic Section of Rocks of East-Central J\,finnesota ........... 102 20. Classification of Igneous and Related Rocks in Central Minnesota 103 2l. Ages for Rocks of East-Central Minnesota (Including Data from Wisconsin and Michigan) ................................... 104 22. Chemical Analyses of Concretions from the Thomson Formation, Minnesota, and from the Michigamme Formation, Michigan .... 115 2:3. Chemical Analyses of Central Minnesota Granites. . . . . . . . . . . .. 117 24. Composition of Granites and Gneisses Assigned to the Minnesota Valley Granite Series ....................................... 124 25. Chemical Analyses of Morton Quartz ::\Ionzonite Gneiss, Odessa Granite, and Bellingham Granite. . . . . . . . . . . . . . . . . . . . . . . . . . .. 129 26. Modes of Sacred Heart Granite. . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 180 9.7. :Modal Analyses of Rocks from the Cedar Mountain Complex .... 138 9.8. Ages for Rocks from Southwestern Minnesota. . . . . . . . . . . . . . . .. 135 9.9. Classification of the Precambrian Recommended by the Canadian National Committee on Stratigraphical Nomenclature ......... 154 30. K-A Ages for Glauconite from the Franconia and .1ordan Forma- tions ..................................................... 164 8l. Comparison of Major Divisions of Present Classification with Older and Possible Future Divisions of the Precambrian of the Lake Superior Region. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 168 ~3':2. Number and Distribution of Dated Samples ................... 171 ABSTRACT Radioactivity dating of a large numher of igneous and metamorphic rocks by the potassium-argon and the rubidium-strontium methods is the basis for revision of the classification of the Precambrian rocks of Minne- sota. The major divisions of the three-fold classification are made at time boundaries of 2.5 and 1.7 billion years (b.y.), corresponding to the time of two major orogenies, the Algoman and the Penokean, respectively. The eras are referred to as Early, Middle, and Late Precambrian in preference to the older terminology of Earlier, .Medial, and Later Pre- cambrian. The division between the Early and Middle Precambrian is placed at the time of the Algoman rather than of the older Laurentian orogeny. The division between the Middle and Late Precambrian is made on the basis of the Penokean orogeny which resulted in a mountain chain that extended from central Minnesota through Wisconsin into Michigan. The Penokean JVlountains formerly were assigned erroneously to late Keweenawan time. The Early Precambrian rocks are divided into the Ontarian and the Timiskamian systems. The Ontarian rocks include the Keewatin group of Minnesota and the Coutchiching metasediments which underlie the Keewatin greenstones in Ontario. Some of the gneisses in the Giants Range and the Vermilion granite regions of Minnesota probably were derived by metamorphism of ancient sediments that were deposited prior to the great outpouring of basalt flows assigned to the Ely greenstone of the Keewatin group. The Laurentian orogeny, although not closely dated, occurred 2.6 bil- lion or more years ago. Folding was accompanied by intrusion of numerous dikes and plutons of tonalitic to granodioritic composition, the largest of which is the Saganaga granite mass on the lVlinnesota-Ontario boundary. The Ontarian rocks and the Laurentian intrusives are overlain uncon- formably by the Knife Lake group which is correlated with the Seine of Ontario. These metasediments are assigned to the Timiskamian. The sedi- mentary cycle was ended by the Algoman orogeny or mountain building. The Giants Range and the Vermilion batholiths, better called granitic ('omplexes, were formed during the Algoman orogeny. Radioactivity ages for these rocks range from 2.6 to 2.5 b.y. Similar ages were found for the Morton gneiss in the JVIinnesota River Valley. Postkinematic granite plutons, emplaced in the folded Coutchiching metasediments in the Rainy Lake area, Ontario, were mapped by Lawson as Algoman. They are dated at 2.4 b.y. Postkinematic granites in the Minnesota River Valley, notably the Sacred Heart granite, give a similar age. XIX )LX PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY The unconformity at the base of the Middle Precambrian Animikie group of the Mesabi district represents a hiatus of several hundred million years. The Animikie formations, the Pokegama quartzite, the Biwabik iron-formation, and the Virginia argillite were deposited on a relatively stable craton of the Early Precambrian granitic rocks. The strata were tilted and warped in the Penokean and subsequent folding but were not highly deformed. To the south, hO'wever, the iron formations and related sediments in the Cuyuna district and the Thomson formation in east- central Minnesota were tightly folded. Farther south, the McGrath gneiss, dated at 1.7 b.y., is a deeper-seated synkinematic phase of the Penokean orogeny. The St. Cloud Red granite and similar intrusives in Wisconsin and in Michigan, dated at 1.65 b.y., are postkinematic and were intruded following the main period of Penokean folding. Tonalites and granodiorite in the St. Cloud area, dated at 1.8 b.y., may represent an early phase of the Penokean cycle. Similar relationships are found in the Minnesota River Valley where the Montevideo granite gneiss was intruded or folded at approximately l.8 b.y. This event was followed by the intrusion of a system of basaltic dikes and then by emplacement of a small granite pluton, dated at l.7 b.y. The Late Precambrian rocks include the Puckwunge formation, the North Shore volcanic group, the late Keweenawan Fond du Lac and Hinckley sandstones, the Duluth complex, and similar intrusive rocks of the North Shore region. A hiatus of several hundred million years is rep- resented in the unconformity at the base of the Puckwunge formation in the Duluth area. The Duluth complex, formed by a number of successive intrusions, was emplaced at approximately l.1 b.y. In the southwestern part of the state, the Sioux formation of Late Precambrian age was folded at about l.2 b.y. Ages determined by the potassium-argon and by the rubidium-stron- tium methods generally are in good agreement; however, the apparent age obtained by either method may reflect metamorphism subsequent to the original crystallization. The study of the possible loss or gain of parent and daughter nuclides in the minerals and also in the whole rock under diverse geologic conditions is an important area for future investigations. The geologic names formerly used in Minnesota and in the Lake Supe- rior region have been retained. It is recognized that some of these terms no longer possess their original usefulness; nevertheless, the rejection of the old terms, as well as the introduction of new names, is inadvisable for the present. The major structural disturbances recorded in the Precambrian rocks of J\Iinnesota and adjacent areas can be correlated tentatively with oro- genic cycles in other parts of the world. It is not likely that the world- wide occurrence of radioactivity ages in the ranges: 2.6-2.5 b.y.; l.8-l.6 b.v.; and l.1-0.9 b.y. is due to chance. The geosynclinal accumulation of sediments and the mountain building that follows cannot be explained ABSTRACT xxi through any agency or force acting entirely within the crust of the earth. Deeper-seated or subcrustal phenomena are involved, and with such hy- potheses the case for world-wide orogenies is improved. Equally improved are the prospects that eventually a world-wide time scale for the Pre- eambrian rocks will be evolved. THE PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY OF MINNESOTA I. INTRODUCTIOi'\ The State of Minnesota, situated on the southern edge of the Canadian Shield (Fig. 1), is in large part underlain by rocks of Precambrian age. In the northern counties the Precambrian rocks are exposed at the surface or are encountered at relatively shallow depths beneath the cover of the glacial deposits. They contain the iron ore deposits of the Mesabi and Ver- milion districts. A large part of the east-central counties also is underlain by Precambrian metamorphic and igneous rocks. Locally, granites of it variety of color and texture are quarried, and St. Cloud is the center of the stone fabrication industry of the state. In the southern counties the Pre- cambrian basement is overlain by sedimentary formations of early Pale- ozoic and Cretaceous age. These strata have been penetrated by lUany water wells which are bottomed in the Precambrian basement. There are numerous smaH exposures of the Precambrian rocks in the southw'estern part of the state, particularly in the Minnesota River Valley where out- crops of granite and gneiss are the sites of numerous quarries. Because of their extent, varied character, and economic significance, the Precambrian rocks of Minnesota have been the subject of geologic investi- gations for many years. Organized geologic work commenced in 187Q with the appointment of N. H. Winchell as State Geologist in charge of the Min- nesota Geological and Natural History Survey. Twenty-four annual re- ports and six large volumes of the final report were published under his direction. Volume 6, a geologic atlas with synoptical descriptions, was pub- lished in 1901, and from that date to approximately 1911, the Minnesota Survey was relatively inactive. Fortunately, during that period geologists of the United States Geological Survey, coming into the state with a broad and varied experience in the Precambrian of other areas, carried on exten- sive field investigations which led to the publication of three important monographs (Clements, 1903; Leith, 1903; Van Hise and Leith, 1911). In 1911 W. H. Emmons became director of the reactivated Minnesota Geological Survey. Vigorous field investigations were renewed, and a num- ber of bulletins were published. Many of the investigations dealt with problems of Precambrian geology and were more or less under the guidance of the late Professor F. F. Grout. This work has been summarized in a paper by Grout, Gruner, Schwartz, and Thiel (1951) that reviews not only the contributions made by the older Minnesota Geological and Natural History Survey and by the present Survey but also the contributions of the United States Geological Survey and the numerous papers by other geologists, many of whom were employed by mining companies. l\Iore re- cent contributions to the Precambrian of Minnesota are compiled in a 2 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY 70· 50 4 0 · 9 0· 8 0 · 70· 60· FIGURE 1. - Sketch map showing the position of the State of Minnesota on the southern margin of the Canadian Shield. The geologic provinces of the Shield in Canada are a fter Collins, Farquhar , and Russell (1954, p . 18) . The dates indicate the approximate time of a major orogeny in each of the provinces. The Precambrian area in Minnesota, as well as in Wisconsin and Michigan, is the southern extension of the Superior province. As expl ained in the tex t , this area has undergone four periods of orogeny, a t 1.1 , 1.7, 2.5, and 2.6 or more billion yea rs ago . guidebook, The P1'ecambrian of N ortheastern Minnesota, edited by G. M . Schwartz (1956a) . The contributors to the guidebook summarized different phases of the Precambrian geology of northeastern Minnesota and also added much new inform ation . THE PROBLEM As a result of the many investigations briefl y outlined above, the Pre- cambrian succession (T able 1) in Minnesota probably is as well known as that of any region of comparable size. The main obj ective of the present INTRODUCTION g investigation was to date by radioactivity methods as many of the intru- sive and metamorphic rock types as possible and to build the framework of a time scale within which the formations and events as interpreted from geologic studies can bc fittcd. The need for such a study has been antici- pated hy our colleagues in their review of the Precambrian stratigraphy of Minnesota (Grout et al., 1951, p. 1017), and with regard to the methods of correlation of Precambrian rocks they conclude: "]\I(ost valuable arc the ages determined by radioactivity, the characters of the zircons in large in- trusions, the petrographic peculiarities of the iron-formations, and the sequence of beds." Since this statement was written, considerable advances ha ve been made in radioactivity-dating methods, and the potassium-argon and rubidium-strontium methods have greatly extended the possibilities of determining the ages of Precambrian minerals. The plan for a program of radioactivity dating in Minnesota was formu- lated in 1955. Facilities for chemical and mass spectrometric investigations were already available for the most part, and a grant from the National Science Foundation permitted initiation of the project in 1956. The potas- sium-argon method was selected for the early phase of the study, and most of the results here reported were obtained by this method. A second grant from the National Science Foundation for the years 1958 and 1959 per- mitted the addition of the rubidium-strontium dating method, and a large number of the samples were also dated by this technique. A third phase of the investigation, involving uranium-lead dating of zircons was originall~­ planned, but circumstances made it necessary to terminate the project. An important objective of the study is the interpretation of the Pre- cambrian history of Minnesota and adjacent areas. In addition to the col- lection of samples, geologic studies were carried on in a number of critical areas in northern Minnesota and in Ontario, and for this purpose field camps were maintained during the summer months of 1956,1957, and 1958. The work in Ontario was particularly necessary because many of the older Precambrian formations were first studied and defined in Ontario. Shorter field studies involved traverses in the Steeprock Lake and the Kashabowie areas, along the highways from Port Arthur to Kenora and International Falls, across the area of the Knife Lake slates in Minnesota to Saganaga Lake and Northern Light Lake in Ontario, in central and southwestern Minnesota, and in Wisconsin and Michigan. PRESENTATION OF RESULTS The geologic succession and chronology of the Precambrian of Minne- sota are shown in Table Q. Ages have been assigned to many of the rocks, but many uncertainties remain, and areas for further field and laboratory studies are indicated. The three-fold division of Precambrian time (Table 1) is retained, but the terminology of Earlier, Medial and Later for the major subdivisions is replaced by Early, Middle and Late. Precambrian time is divided at 2.5 and 1.7 billion years (109 years or b.y.). Major un- >I>- Phanero- zoic Eon Precam- brian (or Cryptozoic Eon) Eras and Rocks Cenozoic Mesozoic Paleozoic Later T"IlL[, 1. l'REC'.Ai\IBHU'I COLU}\IN IN M!N;'\ESOTA. CIIROC-:OLOGIC AND STRATJ(;HAPHIC SEQUEI'CE (After Groul et al .. 1951) Groups Formations and Members (For detail see Minn. Geo!. Survey Bul!. 29) Other Names Used in Minne!ot& unconformity, great on Wisconsin boundary and east-~-~-~--~-~----~----~-~-~~-- U {Sandstones and } {Hinckley pper: other sediments Fond du lac ( Scattered granites" Intrusives, acidic} Duluth gabbro and basic Beaver Bay Complex, Keweenawan Group iMiddle: i and Logan intrUSIves Flows, tuffs and} sediments L w . I Conglomerate and} o er. l sandstones IKeweenaw Point Volcanics Puckwunge formation Red Clastics * ;-unconformity, may be great on the Cuyuna range,~----------------~-~----~---- ---------------Sioux formation"----------- ------------~--~------~--- {Virginia slate (minor iron formation?) = Rove = I Upper Cuyuna slate' Animikie Group Biwabik iron formation series = Gunflint = Deerwood' Pokegama quartzite -~----~-~-unconformity, great on the Mesabi and Gunfiintranges -------------------~-----~--­ Algoman batholithic intrusives, orogeny and erosion Medial Central and southwestern Minnesota granites' (slate Knife Lake Group ~aywack.e (about 18 members) l~~:~~~~lr~~ebeds tuffs, lavas and intrusives Thomson" (Carlton)' Emily' Agawa* Ogishke* ~-----------unconformity, great west of Saganaga Lake------~--------~--------------~-~---~-----(I :r~-~n!~ ~a~e ~a~h~l~h~c ~n~r~si~e~, ~r~g~n~ ~~ ~o~i~ _ Earlier Keewatin volcanics, and Soudan iron formation member. (No Coutchiching recognized in Minnesota) --------I I Ely greenstone • Means uncertain as to place. eo. Era (100 Years) Paleozoic (0.6 b.y.) (Ll b.y.) Late Precambrian (1.7 b.y.) Middle Precambrian (2.5 b.y.) (? b.y.) Early Precambrian TABLE 2. STRATIGRAPHIC SUCCESSION AND GEOCHRONOLOGY OF THE PRECAl\IBRIAN OF :NIINNESOTA (Compare with Table 1) Period-System Major Sequence Formation Orogeny Cambrian Keweenawan Huronian Timiskamian Ontarian ........ Unconformity .. Hinckley sandstone Fond du Lac sandstone ............................ Unconformity ... North Shore volcanic Undivided group Puckwunge Grenville ........... Unconformity ........... . Sioux quartzite (?) ................ Unconformity ............ Penokean Virginia argillite = Rove = Thomson Animikie group Biwabik iron-formation = Gunflint Pokegama quartzite ............... Unconformity .. . Algoman Knife Lake group Undivided ............ Unconformity. . ..... Laurentian Sondan iron-formation Keewatin group Ely greenstone Coutchiching (?) Undivided Older rocks In trusi ve Rocks Duluth complex, sills at Duluth. Beaver Bay complex, Logan intrusives Granite: St. Cloud Red, Rockville (?) granite at Granite Falls, Bellingham (?) Gneiss: McGrath, Montevideo (?) Tonalites: St. Cloud Gray, Warman, Hillman Freedhem, Montevideo Granite: Gold Island, Giants Range, Sacred Heart. Fort Ridgely (?) Gneiss: Giants Range, Vermilion. Morton Saganaga granite, Grassy Island tonalite U) 6 PRECA,MBRIAN GEOLOGY AND GEOCHRONOLOGY conformities are indicated at these dates in Table 2, but it is not implied that these unconformities were formed precisely at these times. The major change from the earlier classification of the Minnesota Geo- logical Survey arises from the recognition of the Penokean orogeny which followed deposition of the Animikie group. The expression Pellokean orof/- cny apparently was first used by Blackwelder (1914) for a period of fold- ing and mountain building which he assigned to post-Keweenawan, that is, at the close of Precambrian time. The Penokean Mountains, as will bc explained in a later section of this report (pp. 118-20), extended from Min- nesota through Wisconsin to Michigan, and possibly connected by way of the ~Iistassini district with a mountain range which formerly existed in the site of the Labrador trough (Fig. 1). Failure to recognize a post- Huronian orogeny has made many correlation problems that have been the subject of papers, as for example, one by Cooke (1926) with the inter- esting title, "Between the Archean and the Keweenawan is the Huronian." Cooke used the name Keweenawan ~lountain range for the Penokean 'lVIountains of Blackwelder. The present work shows that the Penokean Mountains actually are much older than had been previously thought. The term is derived from the Penokee Range in northern Wisconsin. The major orogeny which involved folding of the Knife Lake group and intrusion of the Algoman granites marks the transition from Early to Mid- dle Precambrian time, whereas in earlier classifications (Table 1) the divi- sion was made on the basis of the Laurentian orogeny. In the present classi- fication the Laurentian orogeny is subordinated and is placed within Early Precambrian time. Metamorphism at the time of the Algoman orogeny apparently was so severe that the biotite in many of the older rocks was recrystallized, and the ages determined for biotites from pre-Algoman rocks are essentially the same as those for the Algoman, 2.5-2.6 b.y. Fur- ther discussion of the results is deferred to later sections of this report. The analytical methods used in this investigation are presented in Chap- ter 2. These methods are largely based on procedures developed in other laboratories, and some of the important contributions of these laboratories are reviewed. It is hoped that the detailed descriptions of the methods will give geologists a better understanding of some of the technical problems and also will be useful to others engaged in similar studies. Chapter 2 also includes a comparison of the results by the potassium-argon and rubidium- strontium methods and a discussion of some of the problems in evaluating radioactivity ages. The project, as first formulated and as carried out, is fundamentally a geologic one. The principal Precambrian rock units described in earlier geologic investigations are reviewed, new field and laboratory observations are presented, and the geologic history is interpreted with the aid of the K-A and Rb-Sr ages. The geologic results are presented in three chapters on a geographic basis that is more or less dictated by the distribution of outcrops -in l\linnesota. Chapter 3 deals with the data from northern Min- INTRODUCTIOK 7 nesota and from Canada. Chapter 4 discusses the results for east-central Minnesota and the data from Wisconsin and Michigan. Chapter .5 gives the results for southwestern Minnesota, obtained chiefly for the rocks of the Minnesota River Valley. The general problems of correlation and development of a quantitative time scale are considered in Chapter 6. The locations and brief descriptions of the dated samples, for the most part material not contained in the text, are included in the Appendix. The counties of Minnesota are shown in Figure 37 as an aid to reading and locating samples. In the Appendix and throughout the text, references to public-land divisions in Minnesota for purposes of location are abbrevi- ated, as, for example, "Sec. 22, T. 6.5 N., R. 5 W." to Sec. 22:6.5-5. Descrip- tions with an east range in Minnesota and those outside of the state are given in the full form. 2. METHODS GENERAL STATEMENT Several naturally occurring radioactive nuclides are useful in determin- ing geologic time. Under different geologic environments an age deter- mined from the decay of one radioactive nuclide may be more significant than that from another, but in general the following requirements must be met: 1. The rate of decay of the radioactive nuclide must be accurately known. The half-lives of many naturally occurring radioactive isotopes are of the order of billions of years, and the energy released by the decay of the nu- clides is very small, making the determination of the rates of decay dif- ficult. 2. With the occasional exception of the uranium-lead methods, the tech- niques assume that the mineral or rock is a closed system with respect to both the parent and the daughter nuclides. 3. Appreciable amounts of the daughter nuclide must not be present at the time of formation of the mineral or material to be dated. 4. The isotopic composition of the parent element must have been the same throughout the upper lithosphere. If these requirements are fulfilled, the computed age should correspond with the time elapsed since the sample was last an open system with re- spect to both the parent and daughter nuclides. The age is obtained from the fundamental relationship: Total amount of decay product Age=----------~--~~--~~----~~ A verage rate of production of decay product POTASSIUM-ARGON METHOD DEVELOPMENT At the turn of the century, Thomson (1905) first observed that normal potassium emits weak radiation in the form of ,B-particles. In addition to this activity, Kohlhorster (1930) detected y-emission from potassium. The source of this radioactivity was suggested to be the presence of a radio- active potassium isotope of mass 40 (Klemperer, 1935 and Newman and Walke, 1935). The presence of K40 in natural potassium was verified by Nier (1935, 1936). In view of the dual activity of K40 and the high abun- dance of A 40 in the atmosphere, Weizsiicker (1937) suggested that K40 de- cays to A 40 as well as to Ca40 and that old potassium minerals should con- tain radiogenic A 40. Thompson and Rowlands (1943) subsequently proved that the radioactive decay of K40 involved the production of A 40, and 8 METHODS 9 Aldrich and Nier (1948a) found old potassium minerals to contain appre- ciable quantities of radiogenic argon. Aldrich and Wetherill (1958) have reviewed the radioactivity-dating methods which are useful in determining geologic time and have summa- rized the determinations by various investigators of the specific gamma and beta activities of natural potassium. From the values accepted by Aldrich and Wetherill (1958, p. ~64-~65) and the value of 0.0118 atom per cent for the relative abundance of K40, the computed decay constants used in the present investigation follow: Af3 = 4.76 X 10-10 / yr. A. = 0.589 X 10-10 / yr. The equation used to calculate the A 40/ K40 age of a rock or mineral is - 1 [A40 [A. + Ai3] ] t - A. + Af3 In K40 11.. + 1 . Converting to the common logarithm and using the constants given above, the equation becomes t = 4.31 x 109 log [~:: (9.08) + 1 J. Where duplicate determinations are reported in the tables, the A 40/ K40 values have been averaged and the age computed. SAMPLE PREPARATION Rocks containing fresh, unaltered micas furnish the best material for K-A dating. If the grain size of the potassium mineral is larger than ~OO mesh, a separation is usually made. The rock sample is crushed to less than 40 mesh, and usually the 40-60 mesh fraction supplies adequate material. Most of the biotite separations are made by passing the fraction re- peatedly through a Franz Isodynamic separator until a clean separate is obtained. If the biotite is still contaminated by relatively large amounts of hornblende and pyroxene, the impure material is crushed under a heavy machined steel roller to break up the more brittle hornblende and py- roxene. The biotite is much less disaggregated by this process and is con- centrated by sieving. Repeated treatment effectively removes all but small amounts of the brittle impurities. Bromoform and alcohol mixtures of adjusted specific gravity are used to separate potassium feldspar from the nonmagnetic fraction of the sam- ple. In the preparation of very fine-grained material for which mineral sepa- rations are not practicable, the rock sample is crushed and screened, and the 40-60 mesh fraction is taken for analysis. The mineral samples are gen- erally left as coarse as possible to minimize argon loss during the argon 10 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY extraction procedure, although it has been shown (Goldich, Baadsgaard. and Nier, 1957) that the fine-grained slates of the Thomson formation in lVIinnesota show little argon loss compared to coarse mica in the same for- mation. Portions of the sample to be used for analysis are taken by coning and quartering only after the entire sample has been thoroughly mixed by rolling on hard-surfaced paper. DETERMINATION OF POTASSIUM Experience has shown that the accurate determination of potassium is not the simple matter often supposed. Potassium may be determined by a number of methods such as isotope dilution, X-ray fluorescence, flame photometry and several wet chemical methods. In this work a combina- tion of flame spectrophotometric and wet chemical methods was used. J. Lawrence Smith lJ;1 ethod. Fusion of a 0.5 gm. sample of a finely ground potassium mineral or rock with pure, low-alkali calcium carbonate and A.R. ammonium chloride is carried out in standard-form J. Lawrence Smith crucibles. After cooling, the sintered mass is slaked with water, washed into a beaker, boiled with water and carefully filtered by decanting the supernatant liquid. The boiling leach is repeated three or more times after which the residue is transferred to the filter and thoroughly washed with hot water. The hot filtrate is treated with aqueous ammonia and am- monium carbonate to precipitate calcium, and the calcium carbonate fil- tered off. If the sample contains more than 5 per cent K 20, the residue and the precipitated calcium carbonate are dried, mixed with 0.5 gm. of am- monium chloride, and the entire fusion and leaching process repeated. The combined filtrates are evaporated to dryness in a large fused-silica dish, and the ammonium salts removed by careful heating. The residue is dis- solved in water, treated with a few drops of aqueous ammonia and ammo- nium carbonate, filtered into a platinum dish through a small filter, and the dish and filter washed with hot water. After evaporation of the filtrate to dryness and removal of the ammonium salts, this step is repeated using a weighed platinum dish. When constant weight of the alkali chlorides has been attained through careful heating, a conventional chloroplatinate separation for potassium is made. The separated potassium (plus Rb and Cs) and the sodium (plus most of the Li) fractions are treated with ammonium formate to reduce the chloroplatinates, evaporated to dryness, leached with water and fil- tered. The filtrate is evaporated to dryness, treated with a little sulfuric acid, evaporated and ignited. The alkali sulfate fractions are taken up in water, diluted to 50 mI., and flame spectrophotometric determinations of the K in the sodium fraction and of the Rb, Cs and N a in the potassium fraction are carried out. Blank. Available A. R. "low-alkali" grades of calcium carbonate gen- erally yield a blank of from 0.01 to 0.10 per cent K~O on a half-gram sample 2VIETHODS 11 basis. Careful blanks are run periodically exactly as the analyses, except that a pair of purified alkali chloride blanks are combined and are analyzed with a flame spectrophotometer. With carefully purified calcium carbonate it is possible to reduce the blank somewhat, but rarely to the point where it may be neglected. Thus it is desirable to obtain the best grade of calcium carbonate available. The blank in the present work ranged from 0.01 to 0.06 per cent K 20 (to be subtracted from the determined K 20). After tak- ing into account the blank and correction data, the K~O content of the sample is calculated. Rapid Flame ]/1 ethod. A rapid flame spectrophotometric procedure is used with materials containing less than 3 per cent K 20. A 0.1 to 0.5 gm. sample is decomposed in a platinum dish with hydrofluoric acid and a few drops of sulfuric acid. A solution of magnesium chloride containing the equivalent of one-half the sample weight of MgO is added either before or after the hydrofluoric acid decomposition. After evaporating and fuming off the excess sulfuric acid, the dish is \vashed down, and the residue is broken up with a stout platinum wire. A few drops of sulfuric acid are added, and the evaporating and fuming operation is repeated. The second evaporation is necessary to ensure complete removal of chloride. Following the expulsion of the excess sulfuric acid, the platinum dish is heated carefully, first over a Tirrill, and then over a lHeker burner for 3 to 5 minutes. The alkali sulfates are leached with water and filtered into a volumetric flask which is made up to the mark. Potassium is determined with a flame spectrophotometer, using standards containing 500 ppm. of MgO. Some comparative results are given in Table 3. Precision and Acc·uracy. The J. Lawrence Smith method, the basic pro- cedure in the determination of K 20 for most of the samples, is an extraction method. A consideration of the various steps suggests that if a bias exists that cannot be detected by replicate determinations, the results are some- what low rather than high. High results might be caused by the introduc- tion of potassium from the reagents which is not properly corrected by the blank or by failure to remove all the NH4Cl prior to the conversion to the chloroplatinates. Low results would be caused by failure to extract all the potassium from the mineral and from the calcium carbonate precipitates and by loss of potassium through volatilization during the ignition and in subsequent heating steps to obtain the purified alkali chlorides. Skill and experience are important factors. Replicate determinations made in the Minnesota Rock Analysis Labora- tory over a number of years and throughout the present project rarely dif- fer by more than one per cent relative of the K 20 content in the range of 1 to 14 per cent K 20. The variation in the potassium values obtained by different analysts and between replicate determinations by the same ana- lyst is of the same order of magnitude (Table 4). The accuracy of the potas- sium determinations may be partially assessed by the determinations made on the standard granite, G-l. 12 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY TABLE 3. COMPARISON OF RESULTS FOR K 20 BY RAPID FLAME SPECTROPHOTOMETRIC AND GRAVIMETRIC l\1ETHODS Sample No. KA-238 KA-277 KA-278 KA-39 KA-40 ...... . R-2339 ......... . (C. O. Ingamells, analyst) Weight Per Cent Flame Gravimetric 7.10 5.23 6.85 5.06 4.56 3.44 7.13 5.32 6.68 .5.16 4.56 3.46 TABLE 4. REPLICATE DETERMINATIONS OF K 20 ON GRANITE G-1 Source or Analyst Weight Per Cent Average for 24 laboratories (Fairbairn et al., 1951, p. 37) ...... 5.51 Average recomputed by Fairbairn (1953, p. 146) .............. 5.42 Eileen H. Oslund, Rock Analysis Laboratory (Goldich and Oslund, 1956, p. 813) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.49 S. S. Goldich ...... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.54 S. S. Goldich and Eileen H. Oslund .......................... 5.52 R. A. Burwash (average of 4) ............................... 5.54 Eileen H. Oslund & Halfdan Baadsgaard ..................... 5.52 Average for Rock Analysis Laboratory ....................... 5.52 P. L. D. Elmore, U.S. Geological Survey * ................... 5.51 W. W. Brannock, U.S. Geological Survey * .................... 5.54 * Personal communication from 'V. T. Pecora. May 1959. TABLE 5. REPLICATE DETERJ\IINATIONS OF KoO, Rb,O. AND C&,O ON LEPIDOLITES Weight Per Cent Locali ty and Analysts K 20 Rb,O Cs20 Bikita, Southern Rhodesia Halfdan Baadsgaard & S. S. Goldich ...................... 8.79 C. O. Ingamells & Eileen H. Oslund . . . . . .......... 8.78 Portland, Connecticut Halfdan Baadsgaard & S. S. Goldich ....... . ..... 10.58 C. O. Ingamells & Eileen H. Oslund. . . . . . . .. . ............ 10.48 3.67 3.7.5 0.79 0.80 0.25 0.29 0.32 0.32 In certain samples, such as lepidolites which contain relatively large amounts of all the alkali metals, flame spectrophotometric determinations must be carried out with greater than usual care. The combined gravimetric and flame spectrophotometric method gives good precision for potassium in lepidolites (Table 5). With such complex samples, and with most micas, it has been found necessary to employ a double fusion to ensure complete recovery of the potassium. With samples containing 10 to 14 per cent K 20, 1 to 2 per cent of the KzO content may be retained in the calcium carbonate precipitate and residue unless extremely efficient fusion and extra leach- ings are employed. METHODS 13 DETERMINATION OF ARGON Extraction. The argon extraction system employed in this project is a flux fusion system essentially the same as that described by Wasserburg (1954) and by Wetherill, Tilton, Davis and Aldrich (1956). During the course of the project several modifications were made (Fig. 2A and Fig. 2B). The major change consisted in relocating the magnesium perchlorate and copper oxide traps in the fusion train to remove hydrogen and water as fast as they are generated. The cold trap and tungsten filament were eliminated and titanium sponge metal replaced pyrophoric uranium as the chemical getter. In addition, a standard-taper glass joint was intro- duced just below the sample suspension to eliminate glass blowing. The seal proved to be tight and trouble-free. Mercury cutoffs separate the system into three parts: the pumping system, the fusion train, and the purification train. The pumping system consists of a standard mechanical forepump in series with a two-stage mercury diffusion pump. This system permits rea- sonably rapid evacuation to the point where normal argon contamination is low and the outgassing of the sample, flux, and purification reagents is thorough. Blank runs on the train yield 1-2 X 10-7 cc. STP of A40 and less than 10-9 cc. STP of A36 and A3s. This blank is obtained without a sample, however, and it is usual for the argon contamination from the sample to contribute more than this amount of air argon. For old rocks and minerals, a blank of 10-6 cc. STP of A 40 is usually still small; a correction is applied through the A 36 measurement in any case. To follow outgassing of the sys- tem, as well as subsequent gas transfer and purification, thermocouple pres- sure gauges are situated between the forepump and the diffusion pump and on the purification train. The fusion train consists of a nickel crucible, a sample-mounting device, an A 38 spike, a magnesium perchlorate trap, and a copper oxide trap. The sample to be analyzed is suspended from a steel rod about one foot above the 1 X 12 inch nickel crucible (Figs. 2A and 2B). The crucible is normally charged with 30 grams of sodium hydroxide pellets (low in carbonate) and may be heated to 750 0 C. by means of a removable resistance furnace. The A3S spike is a break-seal tube previously filled with a known amount of A3s, usually 2-5 X 10-5 cc. STP. The magnesium perchlorate serves to take up water from the fusion and is outgassed at 220 0 C. but is used at room tem- perature. The copper oxide trap oxidizes hydrogen to water and is out- gassed and operated at approximately 450 0 C. The purification train contains a titanium-sponge getter and an acti- vated charcoal trap in a break-seal tube. The titanium sponge metal is contained in a stainless steel tube with a Kovar-to-glass seal and is out- gassed and operated at approximately 10000 C., but may be opened to air when cold. The activated charcoal trap is outgassed at 400 0 C. and is used to collect the purified argon at liquid nitrogen temperature. The argon extraction, purification, and collection are carried out as fol- TO ~3e SPIKE FILAMENT TITANIUM GETTER CHARCOAL TRAP TO Hg RESERVOIR CRUCI BLE A FIGURE 2A. - Early a.rgon extraction system with which a pyrophoric uranium ge LLer al so was used in place or the titanium getter shown. TO SAMPLE HOLDER ""A'" SPIKE NICKEL CRUCIBLE ~ NoOH -GAUGE TITANIUM GETTER TO Hg RESERVOIR FIGURE 2,B . - La ler. mod ifi ed argo n ex tl'acLion sysLem. 14 CHARCOAL TRAP B METHODS 15 lows: a weighed sample (1-3 gm.) in a thin glass ampule is suspended from the steel rod extended above the crucible, and an A38 spike, a charcoal break-seal trap and a clean nickel crucible containing fresh sodium hy- droxide flux are attached to the system. The system is evacuated, leak tested, and all the heaters and furnaces except that for the titanium getter are turned on to their outgassing temperatures. The system is allowed to outgas for at least twelve hours, usually overnight. When the outgassing is completed, the flux is allowed to cool, the mer- cury cut-off to the fusion train is raised, and the sample is dropped onto the solidified flux by withdrawing the steel rod with a magnet. The mag- nesium perchlorate is allowed to cool, and the fusion is started by slowly heating the flux to its melting temperature, about 320 0 C. If the fusion proceeds satisfactorily, the A38 spike is introduced by shattering the break- seal tip with a steel ball or slug. The furnace for the titanium getter is turned on, and the outgassing of the purification train is continued. After a fusion of four hours at 750 0 C., the sample and glass ampule have been completely dissolved in the flux and most of the water and hj'drogen elimi- nated. On completion of the fusion, the titanium getter is cooled, and the puri- fication train is sealed off from the pumps by raising the second mercury seal. The first mercury seal between the fusion and purification trains is then dropped to begin the final clean-up of the gas. If the pressure indi- cated by the thermocouple gauge is about 10-2 mm. Hg or less, the titanium getter is heated to 1000 0 C., operated until a constant pressure is attained, and allowed to cool slowly. The heater is now removed from the charcoal trap, and the argon trapped for 20 minutes using liquid nitrogen. The sam- ple is finally sealed off from the system and is ready for mass spectrometric analysis. A 38 Spikes. The A3S spikes are prepared in sets of approximately 36 break-seal tubes. The volumes of the individual tubes are determined, and they are filled with spike A3s on a manifold similar to that described by Wasserburg and Hayden (1955). The A 38 was obtained from the Oak Ridge National Laboratories and contains some A40 and A36 which were deter- mined for nearly all of the spike sets. An average of 15 determinations gave the following argon isotope ratios: A40 / A3S = 0.238± .003 and A36 / A38 = 0.00122± .00001. The spikes are calibrated by two methods. In the capil- lary pipette method (Wasserburg and Hayden, 1955), the spike argon is mixed with a measured amount of normal air argon. The A3S / A38 ratio is used to calculate the amount of A3s present. Two or more spikes of each set are calibrated by this method. The second method of spike calibration employs the multiple expansion of atmospheric argon through a system of known volumes. The apparatus is shown schematically in Figure 3. Valves A and C are double valves and have standard volumes between the seats. Volume A is filled with argon from the tank argon bulb to a pressure which may be measured on the 16 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY TANK ARGON MANOMETER VALVE A EXPANSION VOLUME TO DIFFUSION PUMP VALVE C CHARCOAL TRAP SPIKE A38 FIGURE 3. - Manifold for calibrating A38 spikes. Volumes A and Care 0.504 and 1.002 cc., respectively. Expansion volume B is 17.30 cc. mercury manometer. The gas in volume A is then expanded into volumes Band C. The amount of gas in volume C is further expanded into the right- hand side of the system. The break-seal on the spike is broken, and all the argon in this part of the system is trapped on the charcoal using liquid nitrogen. The quantity of argon delivered by the system is 1.048 X 10-2 PIT. cc. STP where P is the pressure measured on the Hg manometer in cm. Hg and Tr is the room temperature. P is adjusted so that when making the isotopic determination the intensities of the A36, A38 and A40 peaks all fall in the measuring range of the mass spectrometer. Thus a calibration of A 38 can be made simultaneously against A 40 and A 36 of the atmospheric argon. The agreement between the A36 and A40 calibrations is always within 1 per cent. The agreement between two determinations on each spike set is about 1 per cent and between the capillary pipette method and the ex- pansion volume method, about 2 per cent. An average difference of 1.2 per cent between eight spike sets was found. Isotopic Measurements . Most of the argon analyses were made with a 60 ° Nier-type (Nier, 1947) mass spectrometer. The manifolding system used with the instrument is adapted so that samples can be run either by the conventional dynamic technique or by a recirculating technique (Al- drich and Nier, 1948b). A schematic diagram of the mass spectrometer and associated equipment is shown in Figure 4. METHODS AMPLIFIER RECORDER FORE PUMP- I VARIABLE VOLTAGE SUPPLY TO OIL DIFFUSION PUMP 17 FIGURE 4 . - Schematic diagram of argon mass spec trometer tube and gas-handling manifold . Radius of curvature of central ion beam in magnetic field , 6 inches; definin g slits 5" &', and 5, have widths of 0.01 , 0.02, and 0.05 inches, respectively. Ion accelera ting voltage approxi- mately 1800 volts. Ionizing electron beam of approximately 125 microamperes collima ted by magnetic field of approxima tely 200 gauss perpend icula r to diagram in source region . In the dynamic method the gas to be analyzed flows continuously into the mass spectrometer tube through the molecular leak and gas-inlet line at the rate of approximately 10-6 cc. STP / sec. It is pumped away continu- ously by the two mercury diffusion pumps in series with the mechanical forepump. The pressure in the ion source is such that the argon current for A~o, using a tmospheric argon, is of th e order of magnitude of 10-10 am- peres. After passing through slit S3, the resolved ion currents impinge on the ion collector and are measured by the in verse feedback electrometer tube amplifier . Focusing electrodes in the ion source are adjusted so th at in the mass range under consideration the ion current is independent of ion accelerating voltage. Mass spectra are obtained by varying the ion accelerating voltage with a motor-driven potentiometer in the power sup- ply furnishing this voltage, and observing the output voltage of the elec- trometer tube amplifier with a strip-chart recording potentiometer . Figure 5 is a photograph of the mass spectrometer tube and magnet. The ion source is near the bottom center . The gas manifold is behind and around the box housing the valves toward the right of the photograph . The magnet is mounted on a carriage and can be easily withdrawn from the mass spec- trometer tube. A box furnace can be lowered over the entire all -metal mass 18 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY spectrometer tube. The tube it elf is fabricated from stainles steel com- ponents which are welded together. Electrical leads into the tube are through Kovar-to-glass seals . The ion source and collector are mounted on removable flanges, and flange seals are made with copper gaskets. The high vacuum constru ction permits baking of the entire mass spectrometer tube at a temperature of 300 0 C. Occasional bakings reduce residual ion F IGURE 5 . - Pho tograph of mass spectrometer tube, showing ma.in mag~et , !on source mag· ne t, ion source and collec tor assemblies, and gas manifold. DiffusIOn pumps and traps are below table top . METHODS 19 peaks to a level negligible compared with the currents associated with sam- ples under investigation. The great majority of t he samples in the present invest igation were r un dynamically. An extracted argon sample is admitted to the manifo ld through stopcock B. By appropriate setting of the stopcocks, the sample is bled through the pinhole molecular leak at a rate such that the argon peak height decays approximately 1.5 per cent per minute when the ex- pansion volume is 500 cc. Experience has shown that a scan of ten peaks is sufficient to give an accuracy of 1 per cent in the A40 / A 3B ratio . During a run, the ion peaks decay about 10 per cent. Figure 6 shows a typical mass spectrum obtained when the argon extrac ted from a Precambrian biotit e, spiked with A 38 , is run dynamically. Samples are generally analyzed in groups of six or more. The perform- >- l- (/) Z W I- Z :.:: - I --- I T --- - - - - .- T T I I I 4~ I A 100 L ~ - . __ - r- - --- .--- ~ ----w ---- -- - --- - - . - a.. z Q I,r-"! II U I~ In lin In 8765432 0 ---INCREASING T IME (MINUTES) FI GURE 6. - Strip chart of argon from sample KA-24 which had been spiked with AM tracer. Sample was run by the dynamic method, and 10 successive scans of the 40 to 36 region a re shown . Ion currents at mass 40, 38, and 36 positions are approxima tely 8 x lO-n , 2.5 x 10-", and 5 x 10-" amperes, respec tively. Approximately one half of the A 36 in this sample is due to AM impurity in the spike it self, the remainder to atmospheric argon contamina tion arising in the gas extraction and handling process. The atmospheric argon contamin a tion generally does not contribute more tban one per cent to the A'· peak , and correc tion is made as needed . The peak observed a t the mass 39 position comes from the spike and presumably is due to A'·, as it appears with the same abundance in all spike samples and ca nnot be removed bv purifica tion of the gas. Slanting lines show decay of the A' · and A' B, As drawn , the lines fai l sligh tly below the tops of peaks, becau e a correc tion has been made for amplifier zero readings. 20 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY ance of the instrument is checked with atmospheric argon before and after each group of from three to five samples. A typical afternoon's work might thus consist of running six samples interspersed between three atmospheric argon runs. Such a sequence of runs requires approximately ~% hours. The standard atmospheric argon runs (also called discrimination runs) are made primarily to check the mass discrimination characteristics of the mass spectrometer. If the A40 / A36 ratio measured for atmospheric argon deviates from ~95.5 (Nier, 1950), a mass discrimination correction is ap- plied to the data. In practice the A 40 / A 36 measured for atmospheric argon seldom differed by more than ~ per cent from the accepted value. Before and after each run a scan is made of the spectrum over the mass 36 to 40 region. Residual peaks in the instrument are generally so small that no 36 peak is detected, and the correction at the 38 position was in- variably less than 1 per cent of the 38 peak measured on samples. The argon content of samples containing less than 10-4 cc. STP A 40 are measured on the mass spectrometer by the recirculation technique to per- mit measurement of the small but important A36 peak necessary for the atmospheric argon correction. The sample tubes are blown onto a glass line between valves 3 and 5 (Fig. 4). This line is evacuated by diffusion pump No. ~ through valve 5. A sample is admitted to the mass spectro- meter source through valves 3 and ~ after breaking the break-seal tip with an iron slug moved by a magnet. Valve 1 is closed during this operation. There is no constriction in the gas inlet line to the ion source, and the gas is pumped out of the mass spectrometer tube by diffusion pump No.!. In- stead of being exhausted by diffusion pump No. ~ and the forepump, how- ever, it is passed through valve 4 (valve 6 is closed), purified by the tita- nium-sponge getter, and readmitted to the ion source through valve ~. After a sample is admitted to the ion source, valve 3 is closed, and the sample input system evacuated through valve 5 while the sample is recir- culated through valves 4 and ~. Thus, the inlet system is prepared for the next sample while the previous one is being analyzed. After 10 to ~o scans of the 40 to 36 mass region have been made, valve 5 is closed, and the sample exhausted through valve 6. The sensitivity gain of the recirculation method over the dynamic method in the geometry employed is a factor of about 50. The peak heights do not decay with time in a recirculated run as they do in a dynamic run, because the quantity of argon in the recirculating system is not changing with time. A residual blank is obtained in the same manner as the recircu- lated run, since the method is sensitive enough to detect desorption of very small amounts of argon from the walls of the mass spectrometer tube. This "memory" effect is measured before each recirculating run and an appro- priate correction made. Even for young (Cretaceous) samples the residual correction for A 38 and A 40 is less than 1 per cent, unless too short a pumping time is allowed between runs. The mass discrimination factor for recircu- lated runs is determined in the same manD~r as for dynamic runs. METHODS ~l Analytical Errors. The probable error in the determination of the A3< contents of a typical spike set was found to be 1.3 per cent. The error in the potassium determination is estimated to be 1-2 per cent and that for the isotope ratio measurement approximately 1 per cent. The probable total error based only on consideration of the estimated errors for single measurements is approximately ±5 per cent. Variations such as those aris- ing from incomplete recovery of argon from the sample, losses during out- gassing, isotope fractionation in gas transfer procedures, and desorption of residual impurities during mass spectrometric analyses have gradually been minimized by improvements and tests ov'er a period of time, but may appear in individual runs to give erratic results. During the course of the investigation two or more determinations of the argon contents of approximately 90 samples were made, and the aver- age difference in the A40 / K40 ratios is about 3.5 per cent. The average age difference is less. Two or more argon determinations for 54, or slightly more than one-third of the samples, are listed in Tables 14, 21, and 28. Fifty per cent of these differ by less than 3 per cent, and 80 per cent by less than 5 per cent. The rlifference in the A 40/ K40 ratios for nine of the samples ranges from 5 to 10 per cent, and for two samples it is greater than 10 per cent. Of the latter, KA-94 (Table 14) is a sample of biotite for which three argon extractions were made, giving A 40/ K40 ratios of 0.292, 0.302, and 0.332. The age computed for the average ratio is 2.50 b.y., and for the ex- tremes, 2.42 and 2.60 b.y. The second sample is a graphitic slate (KA-35, Table 21) which gave A 40/ K40 ratios of 0.140 and 0.166, corresponding to ages of 1.54 and 1.72 b.y. RUBIDIUM-STRONTIUM METHOD DEVELOPMENT Rankama (1954) has summarized the development of the Rb-Sr dating method from the beginning made by Hahn, Strassman and Walling (1937) up to 1953. Since 1953 great strides have been made in the development and application of the rubidium chronometer, Aldrich and Wetherill (1958) have reviewed the status of the Rb-Sr method and the work done in evalu- ating the decay constant. Since this review, Flynn and Glendenin (1959) have completed measurements on the specific ,8-activity of Rb87 using liquid scintillation counting techniques and obtained a value of 47± 1 X 109 years for the half-life. This value yields A{3 = 1.47 X 10-11 / yr. The equation for the calculation of the Rb-Sr age, t = :{3 In [~~:7 + 1] or [ srS7 ] t (years) = 1.56 x 1011 log Rb87 + 1 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY was used to calculate the Rb-Sr ages found in this report. Comparison ages taken from the literature have been recalculated using the Flynn and Glendenin value for 1..(3. In the analytical computation the value of 27.85 atom per cent was used for the relative abundance of RbR7 • SAMPLE PREPARATION In general the mica samples prepared for the K-A determination were used; however, a small portion of the original sample was further purified, mainly by the crushing and sieving technique previously described. The potassium feldspar samples were ground to less than 200 mesh and centri- fuged in heavy liquids to minimize the amount of plagioclase. The sericite, KA-3~8, from Gold Island, Saganaga Lake, was also centrifuged. AN ALYTICAL PROCEDURES Reagents and Apparatus. The water is double distilled, the hydrochloric acid redistilled as constant-boiling acid, and the oxalic acid solution puri- fied by passing through a large ion exchange column. These three reagents are stored in polyethylene containers. The A.R. grades of hydrofluoric, perchloric, and sulfuric acids are used without further treatment. All glass- ware is cleaned with hot nitric acid prior to use. The cation exchange column employed contains 200-400 mesh, medium porosity Dowex 50 W X 8 styrene-type cation exchange resin. After pre- cleaning with nitric acid and hydrochloric acid, the resin is slowly settled in a pyrex tube, 70 cm. in length and 1 cm. in diameter. The tube is fitted with a fritted-glass disc and tapered at the bottom so that liquid cannot collect in greater than drop-size quantity before entering the collection vessel. Decomposition. A sample sufficiently large to give about about 250 fJ. g. Rb is decomposed in a platinum dish with a mixture of hydrofluoric, per- chloric and hydrochloric acids. An amount of Sr86 spike solution is added to make the abundances of Sr86 and of Sr87 in the sample solution of the same order of magnitude. The solution is evaporated to dryness, treated with additional perchloric and hydrofluoric acids and again evaporated to dryness. Biotite is easily decomposed by this procedure, but muscovite is more resistant, and care must be taken to ensure that complete decompo- sition has been achieved. The residue is dissolved in 6 N hydrochloric acid and made up to 50 ml. in a volumetric flask. Rubidium. With the exception of a few unusual samples, a 5 ml. aliquot of the sample solution is transferred to a platinum dish, and an amount of rubidium spike is added which makes the relative abundance of Rb87 about twice that of Rb85 • A few drops of sulfuric acid are added, the solution is evaporated to dryness, the sulfuric acid fumed off and the residue ignited over a Meker burner. The ignited residue is leached with double-distilled water, decanted or filtered off, and the solution evaporated to dryness for isotope analysis on the mass spectrometer. METHODS 23 Strontium. The remainder of the original sample solution is evaporated to about 10 ml. and enough of the radioactive Sr85 isotope added so that strontium may be traced with a scintillation counter in subsequent steps. A little perchloric acid is added and the solution is chilled in an ice bath to precipitate as much of the potassium and rubidium as possible. After filter- ing or centrifuging, oxalic acid is poured into the ion exchange column, and the filtrate or supernate added. Enough oxalic acid is passed through thc column to elute the iron in the sample quantitatively, and 2.5 N hydro- chloric acid is passed through the column to elute the Sr. That portion of the eluate (about 20 ml.) containing the greatest part of the Sr, as deter- mined by testing the collected portions of the eluate with a scintillation counter, is treated with 1-2 drops of perchloric acid and evaporated to dry- ness preparatory to isotopic analysis on the mass spectrometer. Spike Solutions. Strontium carbonate enriched in Sr86 and rubidium chloride enriched in Rb87 were obtained from the Oak Ridge National Laboratories. The salts were dried, weighed and dissolved in double- distilled water to make the primary spike solutions. The solutions are kept in airtight polyethylene bottles to minimize evaporation and change in concentration. * The absolute concentrations of the spike solutions were calibrated b~' isotope dilution using normal rubidium and strontium salts (Davis and Aldrich, 1953). Purified normal rubidium and strontium salts were used to make up standard solutions, and the isotope ratios were checked on the mass spectrometer. Aliquots of the respective normal salt standard were mixed with the spike solution, and the determination of the isotopic ratios on the mass spectrometer enabled the calculation of the concentration of the spike solution. The values obtained were checked against the gravi- metric values. Several calibrations were made, and the relative abundance of the Sr isotopes in the spike solution was found to be as follows: Sr8 4, 0.0197; Srs", 83.44; Sr87 , 2.028; and Sr88 , 14.52 atom per cent. The Sr86 spike solution contained Sr 86, 3.712; Sr8 ., 0.0913; and Sr88 , 0.6609 fLg./ml. The relatiyc abundance of the Rb isotopes in the enriched Rb8 • spike solution was de- termined to be Rb85 , 4.81; and Rb"', 95.18 atom per cent, and the solution contained Rb85 , 0.127 and Rb8 ., 3.002 fLg./ml. Isotopic Measurements. The rubidium and strontium isotope ratio meas- urements are made on a six-inch, 60° Nier-type mass spectrometer. The usual shield and electron beam apparatus are replaced by a shield milled from a solid block of stainless steel with an opening on the side into which interchangeable stainless steel filament blocks can be slipped. Approxi- mately 1 X 30 mil tantalum filaments are spot welded onto Kovar pins which fit into the filament blocks in such a way that the filaments are aligned 1% mm. behind the first slit of the ion source. Conventional J * There is some evidence suggesting polyethylene bottles are not wholly satisfactory and that. solutions should be checked at 3-6 months inten·als. 24 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY plates and collimating slits are used. The grounded collimating slits are 0.005 and 0.010 inches wide, respectively. A 0.025-inch collector slit and conventional feedback electrometer amplifier (10" ohm resistor) follows the magnetic analyzer. Since ions are formed on the tantalum filament by surface ionization, a source magnet is not used. Strontium ions appear and produce a steady ion beam of the order of 10-12 amperes at about 2.3 am- peres filament current. Rubidium ion beams of about 10-11 amperes are ob- tained at about 1 ampere filament current. One or two drops of the extracted strontium perchlorate are placed on a clean filament and loaded into the spectrometer. It was found that the most stable ion currents are obtained after outgassing the filament (with sample) overnight at 1.5 amperes filament current. Any rubidium present is thus baked off the filament. By slowly bringing the filament current to more than 2 amperes over a period of several hours, strontium peaks appear and can be stabilized at a convenient intensity. Twenty to thirty scans of the 86, 87, and 88 mass regions are sufficient to calculate the isotope ratios with a 1 to 2 per cent error. Constant checking of the mass-85 position during a run will indicate rubidium contamination. Whenever an 85 peak appears, the ion source is cooled for an hour or more before the run can be continued, as rubidium could be emitted from hot source parts other than the filament. After the strontium run is completed, the same filament is used for the rubidium determination. One drop of the spiked rubidium sample solution is placed on the filament and loaded into the mass spectrometer. Steady ion beams are obtained after half an hour of outgassing. In order to ensure that there is no cross-contamination between the samples, the ion source parts are cleaned in nitric acid between each run. In addition, separate shield blocks and slits are used specifically for rubidium or strontium determina- tions. This ensures the appearance of only normal rubidium (as measured at mass 85) in strontium runs, enabling a correction to the mass 87 peak to be made in the case of contamination. Analytical Errors. An important unavoidable contribution to the error in the calculated Rb-Sr age lies in the determination of, and correction for, the amounts of normal strontium present in mineral samples. This amount of strontium is usually small, less than 20 ppm., but often it is relatively large compared with the amount of radiogenic Sr87 present, especially in young or Rb-deficient minerals. Gast (1955) has shown that the variations in the isotopic composition of normal strontium are small, so that a good correction may be made for normal strontium most of the time. If the amount of normal strontium is very large compared to the radiogenic Sr87 , the error in the isotopic abundance of Sr87 in normal strontium will affect the calculated Rb-Sr age. In the corrections for normal strontium the ratios Sr87 / Sr88 = 0.0841 and Sr87 / 8r86 = 0.713 were used. These values based on the measurements of White and Cameron (1948) are essentially the ratios recommended by METHODS Gast (personal communication, 1959). Gast has treated the subject of analytical errors in Rb-Sr dating more completely and calculated probable errors in the Sr87 / Rb87 ratios of from 2 to 3.4 per cent for old Precambrian micas. In the present work the mass spectrometric measurements are con- sidered to be accurate to about 2 per cent, and the probable error in the Sr87 / Rb87 ratios is approximately ±5 per cent. Duplicate determinations on nine samples showed a maximum difference of 6 per cent and an average difference of 2.5 per cent. EVALUATION OF RADIOACTIVITY AGES The evaluation of radioactivity ages is a complicated matter, because many factors are involved some of which are poorly understood. The pres- ent discussion touches briefly on analytical problems, decay constants, and some of the geologic problems. ANALYTICAL PROBLEMS The precision or reproducibility of the methods used to determine the relative abundance of parent and daughter nuclides for the calculation of an age usually can be estimated, but the accuracy is much more difficult to appraise. Efforts to determine probable accuracy include comparison of the basic analytical determinations by different methods and by different laboratories and also comparison of the age found for a rock or geologic formation by different dating techniques. A number of samples has been available through the courtesy of the geochronology groups of the Carnegie Institution of Washington, the Massachusetts Institute of Technology, the Lamont Geological Observ- atory, and others. In an earlier paper (Baadsgaard, Goldich, Nier, and Hoffman, 1957) a comparison of the A 40/ K40 determinations on the same minerals made by a number of groups suggested that the reproducibility among these laboratories was of the order of five per cent. Analytical data for some of the interlaboratory check samples have been recently published by Aldrich, Wetherill, Davis, and Tilton (1958), and in the discussion that follows some comparative data from the Carnegie In- stitution of Washington and from the Uninrsity of Minnesota are given (Table 6). The basic analytical data are in parts per million by weight; K40 has been computed on the basis of K40 = 1.21 X 10-4 gm./gm. K; and the ages are computed with the constants used in this report and given in bil- lions (109 ) of years. Hikita QuarTY Lepidoll:te. The A40 / K40 and Sr87 / RIP ratios for lepid- olite from the Bikita quarry in Southern Rhodesia, calculated from the data given by Aldrich, Wetherill, Davis, and Tilton (1958), are in good agreement with the results obtained at Minnesota. Although the ratios and the ages from the two laboratories differ by less than 1 per cent, the analyt- ical determinations made at Minnesota are consistently lower by 2-3 per 26 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY TABLE 6. ANALYTICAL DATA FOR LEPIDOLITE. IVluSCOVITE, AND BIOTITE FROlH INTERLABORATORY CHECK SAMPLES AW K'lO Age Sr"' Rb" Age Lab. ppm. ppm. A'"/K''' b)·. ppm. ppm. Sr"/RI/' b.y. Lepidolite (Southern Rhode~ia) Carnegie Inst. '2.-Hi !l.U.; O.27~ ~.33 356 9370 U.0380 2.63 lini\'. of l'vlinnesota '13~ 8.S1 0.27U 2,32- 3+8 9110 0.0382 :'1.64 Jl1lscovite (North Carolina) Carnegie Inst. U.23'2 10.78 lI.0216 0.330 1.43 ~70 0.00.129 O.S(iO 1.50 216 0.00545 0.370 Uni,'. of Minnesota 0.217 10.18 0.0\113 0.330 Biotite (Te.ws) Carnegie Insl. (granite) ... 0.370 4,44 00832 1.05 1.77 115 0.0154 1.04 l'ni,'. of Minnesota (pegmatite) . O.6R6 fLU) O.ORS;; 1.0(i iJ.SS (;34 0.0156 1.0.'> cent. A possible explanation is that the samples were slightly different, but compensating errors cannot be ruled out. Chestnut Flat 'A-line 'Muscovite. Comparative analytical data for mus- covite from the Chestnut Flat mine, Spruce Pine district, North Carolina, show considerably larger variations than were found for the Bikita lepid- olite. Although the basic analytical data differ by 6 per cent, the result- ing A'o / K41J ratios differ by 1 per cent, and the agreement must be the result of differences in the samples or of compensating errors, Duplicate Rb-Sr determinations in the Carnegie Institution laboratory differ by 3 per cent. Petrick Quarry Biotites. The samples of biotite analyzed in the Carnegie Institution and at the University of Minnesota are not the same and can- not be considered interlaboratory check samples, but the agreement in ages from the two laboratories is noteworthy. The granite and pegmatite of the Petrick quarry in Llano County, Texas, have been described by Gold- ich (1941) and by Barnes, Dawson, and Parkinson (1947), Biotite dated by the Carnegie Institution group is from the granite, but the biotite dated at J\linnesota is from a pegmatite in the granite; hence the analytical data are not comparable. COMPARISON OF K-A AND Rb-Sr AGES The analytical data and the Rb-Sr ages for 35 samples aTe given in Table 7. Four of these samples were analyzed by P. W. Gast in the Lamont Geological Observatory and one by George Edwards in the Shell Develop- ment Company laboratory in HOllston, Texas. The K-A ages, all deter- METHODS 27 mined at Minnesota, for 32 samples of mica are included in Table 7, and the analytical data for these samples are given in tables that follow. Age determinations for miscellaneous samples are presented in Table 8. Figure 7 is a plot of the K-A versus Rb-Sr ages for the 32 samples (Table 7) but also includes five additional samples which will be reported in papers now in press or in preparation. The scattering of the points in Figure 7 may be attributed in part to analytical errors; each point represents a pair of ages for which four analytical determinations are required. The scatter may also reflect the different behavior of the micas with regard to the re- tention of argon and of strontium during metamorphism. The arithmetic mean of the ratio, Rb-Sr age/ K-A age, for the 37 sam- ples is 0.99, with a range of 0.90 to 1.10. The difference between the two ages is less than 5 per cent for 30, or 80 per cent, of the samples and be- tween 5 and 10 per cent for the remainder. Aldrich, Wetherill, Davis, and Tilton (1958) give comparative K-A and Rb-Sr ages for micas from 37 localities. Three of these localities are duplicated in the Minnesota samples, but the others are widely distributed in North America and include six 2.5 / (/) a::: <{ 2.0 w >- o a> 0 W (!) 1.5 <{ / <{ I :.:: 1.0 fJ 1.0 1.5 2.0 2.5 Rb-Sr AGE (109 YEARS) FIGURE 7. - Plot of K-A yersus Rb-Sr ages. Grouping o r the samples re- Aects the ma.jor periods 01" orogeny; Algoman , -'l .5 b.y. , Penokean , -1.7 b.y ., Grenville, -1.0 b.y. Data ill Table 7. TABLE 7. Sr"/Rb" AGE DETERMlNATIONS (Data in parts per million by weight) Rb-Sr Age K-A Age Total Normal KANa. b.y. b.y. Description Rb Rb" *Sr~7 Sr Sr"/Rb" (A) Northern Minnesota 17M" 2.50" 2.4R Rader pegmatite. Northwest Angle ...... 4730 1340 50.4 2.0 0.0376 88B 2.47 2.47 Pegmatite. Northwest Angle ............... l~OO 340 1'2.6 56.3 0.0371 239B 2.47 2.57 Vermilion granite. Crane Lake ...... .593 168 6.'2.5 9.2 0.037'2 4RB 2.51 2.33 Giants Range granite, Mountain Iron ....... 727 206 7.79 10.4 0.0378 237B 2.33 2.53 Giants Range granite, Nashwauk .. 364 103 3.61 7.9 0.0350 328S 2.50 2.43 Pegmatite, Gold Island, Saganaga Lake ..... 274 77.6 2.9'2 34.3 0.0376 3'28F 2.49 Th~ .................................. 470 133 4.98 45.1 0.0374 307B 2.28 2.38 Tonalite, Grassy Island. Rainy Lake ........ 470 133 4.5.5 33.4 0.0342 65B 1.09 1.06 Gabbro. drill hole. Babbitt .............. 470 133 2.17 18.9 0.0163 296B 1.12 Pegmatite in iron formation, Babbitt ... 1890 .535 8.89 12.2 0.0166 (B) Ontario >C 1I2B 2.77 (?) 2.60 Schist, drill hole, Brim'c1iffe Lake ........... 306 86.6 3.61 ILl 0.0417 00 76B 2.50 2.59 Northern Light gneiss. Moose Island. North- ern Light Lake ................. 250 70.9 2.63 26.3 0.0371 253 71.7 2.73 18.5 0.0381 77B 2.59 2.60 Ditto. Savage Bay ........................ 339 95.9 3.74 20.4 0.0390 262B 2.59 2.63 Granite gneiss, English River .............. 487 138 5.38 8.2 0.0390 19lB 2.49 2.52 Coutchiching schist, Rice Bay, Rainy Lake 262 74.3 2.86 3.8 0.0385 276 78.2 2.83 15.2 0.0362 320B 2.45 2.50 Aplite, Rice Bay, Rainy Lake ....... ,530 150 5.51 27.5 0.0368 222B 2.36 2.41 Granite, Ben Island, Rainy Lake ... 678 192 6.81 37.3 0.0355 2'24B 2.34 2.45 Granite, Goose Island, Rainy Lake. 434 123 4.26 13.7 0.0346 424 120 4.25 10.,5 0.03.54 272B 2.38 2.36 Granite gneiss, Emerson Lake ..... 625 177 6.33 7.1 0.0358 (C) Central Minnesota lB 1.72 1.76 Quartz monzonite, Warman .... ........... 427 I'll 3.11 17.0 0.0257 64B 1.69 1.77 Tonalite, Hillman .................... 484 137 3.45 17.3 0.0252 4lB 1.68 1.67 Granite gneiss, Denham ................... 420 119 2.96 48.2 0.0249 406 115 2.90 5.2 0.0252 58B 1.65 1.64 St. Cloud Red Granite .................... 886 251 6.17 8.5 0.0246 164B 1.65 1.71 McGrath gneiss, McGrath .............. 515 146 3.61 10.5 0.0247 w CD TABLE 7 - Continued Rb-Sr Age K-AAge Total KANo. b.y. b.y. Description Rb (D) Sonthwestern Minnesota 14B 2.43' 2.53 Morton gneiss, Morton .................... 748 107B 2.42 2 .. 57 Morton gneiss, NW of Morton ............. 547 13B 2.45 2.43 Sacred Heart granite ...................... 692 696 9B 2.40 2.42 Sacred Heart granite ...................... 745 798 27B 1.73' 1.85 Montevideo granite gneiss, SE of Montevideo 957 1.75 953 54B 1.69' 1.75 Montevideo gneiss, Granite Falls ........... 477 29B 1.65 1.63 Granite, Granite Falls .................... 431 (E) MUicellaneoliS OSD(L) 2.54 2.32 Pegmatite, Bikita quarry, Southern Rhodesia. 32170 128B 1.05 1.06 Pegmatite, Petrick quarry. Llano County. Texas ................................. 2240 OSA(F) 1.62 Pegmatite, Dickinson County, Michigan .... 1020 190 0.73" 0.74 Illite, Siyeh limestone, Belt series, Glacier Na- tional Park, Montana ................... 235 * Radiogenic Sr87. a Letter denotes mineral: (B). biotite; (F), feldspar; (L), lepidolite; (M), muscovite; (S), sericite. b Gast, Kulp, and Long (1958). 'Determination by P. W. Gast (personal communication, 1959). d Determination by George Edwards (Goldich, Baadsgaard, Edwards, and Weaver, 1959). Rb" 212 155 196 197 211 226 271 270 I35 122 9110 634 289 66.6 TABLE 8. A '''/K'" AGE DETERMINATIONS FOR MISCELLANEOUS SAIIlPLES KoO K1- • ~ rJ) /J 0 1.2 1.0 // w // (!) O.B . SERICITE SCHIST, AND SERICITE SAMPLES FROl\1 SOUDAN AREA Si02 · . 49.7~ AbO" 16.76 Fe20, . ~ 1.92 FeO · . 7.33 MgO 7.G2 CaO 9.35 Na,O 3.14 K,O .71 BaO ............ .06 TiO, ... .8!) p,O" · . .O!J MnO . ](j H,O+ 1.57 H,O- .. .0(; CO2 .10 S ....... .04 99.52 <0=8 .01 Total 99.51 2 31.85 \17.93 2.15 :13.38 1.33 .04 .43 3.26 .81 .08 .08 8.1 ~ .48 .04 09.98 99.98 3 ,;9.71 ]6.0.5 .73 l.SI 2.78 4.5.5 .5] 4.JI) .2S . Hi .15 ~.O7 .29 (;.40 !HI.G7 99.(i7 KA-97 U.61 (j'(H KA-lS U.;j] HAl 1. Ely greenstone, M4011, from outcrop, Sec. 9: 61-14. Eileen H. Oslund, ana- lyst (Schwartz and Reid, 1955, p. 299). 2. Altered wall rock. drill hole on 19th level. Soudan mine. H. Baadsgaanl. analyst (Ibid.). 3. Sericitic schist, drill hole 7U8, 175-210 ft., 12th level. Soudan mine (lbid.)~ KA-97. Sericite, cross cnt north of main drift on 27th leye!, Soudan mine, Doris Thaemlitz, analyst. KA-18. Sericite, north wall of Shaft Vein orebody, approximately 25 feet above 27th level, Soudan mine, Eileen H. Oslund, analyst. oS;; in the mesozone." According to Buddington, granites of the mesozone, for the most part, are emplaced at depths ranging from 4 to 10 miles. K-A ages for three of the rocks of the Snowbank stock are given in Table 14, Section C. If the ages are taken without interpretation, they might bc said to indicate intrusions at widely separated times; microgranite por- phyry (KA-229) at 2.3 b.y., a second intrusion metamorphosing thc Knife Lake conglomerate (KA-323) at 1.7 b.y., and finally a granite (KA- 106) at 1.2 b.y. The A 40 /K40 ratio of K-feldspar separa ted from the granite (KA-106), however, is greater than that of the associated biotite (Fig. 8). This clearly indicates that the biotite has been recrystallized and that the determined age is a survival number and not the time of crystallization of the granite. The contact of the transgressive Duluth gabbro is just south of Round Lake (Fig. 14), approximately 1.5 miles south of sample KA-106. The gabbro was intruded at approximately 1 b.y. and has affected the granite. The biotite, being more sensitive than the feldspar, apparently lost all or the greater part of its argon at the time of intrusion of the gab- bro. If the granite is Algoman, as the field data indicate, the feldspar also lost considerable radiogenic argon during this event. The 2.3 b.y. age, ob- j{ i{ Q) "'" o ..J 1{ Duluth gabbro r-+""+l l±........:LJ Gronites ond syeni tes Feldspar porphyry LIIIITIJ Agglomerates and conglomerates f "~'\} " ',"·!AI Xt~ · Co ,,-~,.;:~ Conglamerates Slates, graywackes 10 o /I Geologic boundory Probable faul! v < .; ~ .; < A KA-229 • 2.3 F IGURE 14. - Geologic map of tbe Snowbank Lake area, showing location of dated samples. After Gruner (1941 ). 56 NORTHERN MINNESOTA 57 tained for biotite separated from the microgranite (KA-229), supports the geologic assignment of the Snowbank stock to the Algoman, and it seems likely that the ages are low because of subsequent metamorphism, increasing from south to north and northwest. Saganaga Lake Area. Efforts to separate biotite, suitable for dating, from the Saganaga granite were unsuccessful. Biotite is not abundant and generally is chloritized. Sericite from the pegmatite on Gold Island is dated by the K-A method at 2.43 b.y. and by the Rb-Sr method at 2.50 b.y. Feldspar from the pegmatite gives a similar Rb-Sr age of 2.5 b.y. Two samples from the Northern Light gneiss, collected from Northern Light Lake, east of Saganaga Lake in Ontario (Fig. 13 and PI. 3) were dated by both K-A and Rb-Sr methods, and the age is 2.6 b.y., although one of the Rb-Sr determinations is somewhat lower, 2.5 b.y. The ages by themselves would suggest that the Saganaga granite is younger than 2.6 and older than 2.5 b.y. However, this interpretation may not be justified, as will be shown later in the discussion of the history of Early Precambrian time. Biotite from the large lamprophyre dike in the Saganaga granite is dated at 1.S b.y. (Table 14, Sec. E), and this age indicates activity at a much later time. Whether the age is the time of intrusion of the dike or the time of later alteration is uncertain. It seems possible that the low ages on the samples of the Knife Lake slate may be in some way related: although the age of O.S b.y. for the slate west of Ogishkemuncie Lake re- quires another explanation. Efforts to date the orthoclase, which makes up the veins in the Saganaga granite at Gull Lake, by the Rb-Sr method were unsuccessful because of the large amount of normal strontium in the feldspar. VERMILION GRANITE REGION GENERAL STATEMENT Northwest from the Vermilion district to the Ontario border is a region of high-grade metamorphic and granitic rocks. A coarse-grained, light- pink biotite granite is prominent, and Grout (HJ23, p. 254) named this rock the Vermilion granite from the good exposures on Vermilion Lake and along the Vermilion River. The Vermilion hatholith (PI. 1) measures SO miles in an east-west direction and approximately 30-40 miles llorth- south. This large area, which occupies most of northern St. Louis County and parts of Lake and Koochiching counties, is not very well known. The most detailed paper is that of Grout (1925b), which includes chemical analyses and petrographic descriptions. The Vermilion granite is intru- sive into the Ely greenstone and also into mica schists, which have been correlated with the Knife Lake group of the Vermilion district. On this basis, Grout assigned the Vermilion granite to the Algoman, so that the Algoman granites in Minnesota, including the Giants Range and the Ver- milion granites, far overshadow the older Laurentian Saganaga granite in areal exten t. ;;8 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY T.-\BLE: Itl, CHE.:\[lCAL ANALYSES OF VERi\IILION, GIA~TS RANGE, AND SNOWBANK GRANITES, A;>;D OF GNEISS FROi\I BURNTSlDE LAKE Yennilion Giants Range ~ 3 4 SiO, 71.73 72.06 66.31 71.45 ALO," 14$5 16.0;; ]2.77 13.11 Fe,O" .58 .4(; 1.98 2.44 FeO .. 1.3;; .72 2.66 1.78 }igO .(:~ .07 3.01 .84 CaO ... 1.18 .86 4.77 .65 Na,O 3 .. 58 4.56 5.03 3.03" K,O ~.63 3.54 2.33 4.79b BaO .1 " . 1 '2 .00 .00 TiO, .33 .12 ,49 .13 P,Oo .14 . 00 .13 .04 MnO . 03 .06 .04 .04 H,O+ .. .64 .39 .75 1.24 H,O- .. ... . .20 .0;3 .05 .10 CO, .... lI.d. . 10 n.d. n.d . FeS, .... . .0(; .09 lr. tr . Total ... 100.25 100.24 100.32 99.64 " Figures for Al,(}, include Zr02 and Cr,O.1 reported originally. h Original alkalies reported by Allison (1925) are in error. Snowbank Burntside ij () 69 . .50 68.54 16.86 17.89 .50 1.77 .64 .52 1.45 1.22 1.70 4.M 4.58 .5.14 3.94 1.05 n.d. n.d . .,;3 .20 n.d . II'. n.d. Il.d . .42 .4G .20 .IS n.d. n.d. n.1.97 ALO, 14.88 12.58 12.73b 12.05 18.97 Fe,O, 7.19 4.70 4.36 2.37 1.98 FeO .... 8.14 4.83 3.93 1.09 7.07 MgO .. 2.36 .93 1.66 .51 2.67 CaO .... 7.06 2.07 1.20 .76 8.60 Na,O .... 3.74 3.46 3.45 2.85 3.18 K,O .. 1.68 4.21 3.98 5.40 1.43 BaO ... n.d. .12 . 06 .07 n.d . TiO, 2.15 .82 .69 c)' ._1 1.90 p,O, ...... .82 .15 .11 .05 .41 MnO . . . . . . . .20 .16 .16 .05 .14 H,O+ ............... .87 .54 1.25 .38 1.31 H,O- ....... .21 .32 .20 .09 .17 CO, .. .02 .03 . 02 .16 n.d . S .... n.d. .03 . 0 .. .01 n.d . 99.71 99.90 100.76 99.67 99.79 ... 1- Gabbro (C.dor Mounloin Iyp.) ~ QOQ Morton quartz monlonite gneiss i gabbro oneiss inclusions (goQ) .. Somple locotion 1.8 (K-A AQe, b.y.) FIGURE 30. - Geologic map of the Cedar Mountain area south o f Franklin . SOUTHWESTERN MINNESOTA 133 The country rock of Morton gneiss is in no way exceptional and shows the usual contorted structure. A number of small xenoliths of the gneiss occur in the gabbro and are easily distinguished from the red granite of the core. A modal analysis of the olivine trachybasalt porphyry from the chilled border zone is given in Table 27. Phenocrysts of augite and olivine occur in a fine-grained diabasic groundmass. Orthoclase is interstitial to the laths of plagioclase (An54)' Olivine is largely altered to serpentine and magnetite. Granophyre Gabbro. The gabbro is characterized by well-developed layering. On flat outcrops the banding shows swirl-like structures as though a crystallizing magma were slowly rotating as it rose. On the cliffs the layering appears as vertical, parallel bands. The structure is due to variations in the proportions of light- and dark-colored minerals. The light-colored layers are mostly feldspar laths, whereas the darker ones are chiefly augite, hornblende, biotite, and magnetite. The bands range in width from V8 inch to 6 inches or more. The light-colored layers commonly are wide and commonly stand above the dark bands on weathered surfaces. In thin sections the gabbro shows a very large amount of alteration. The hornblende llnd possibly the biotite may be classed as deuteric, but abun- dant chlorite, epidote, sericite, and magnetite are hydrothermal. The plagioclase (An55) is extensively altered to sericite. An average of four modal analyses is given in Table 27. Granite. The largest outcrops of granite occur near the center of the Cedar Mountain complex. A small patch of granite occurs outside of the gabbro on the southeast side, marginal to a dike-like mass of gabbro that may be an apophysis of the main mass. Small dikes of granite cutting the TABLE 27. MODAL ANALYSES OF ROCKS FROM THE CEDAR MOUNTAIN COMPLEX (Volume per cent; Curtis A. Bury, analyst) 1 2 3 4 5 K-feldspar .............. , ... II 18 47 39 41 Plagioclase " , ............... 50 30 25 18 15 (An,,) (An",,) (An.) (An.) (An.) Quartz ••••••••••••••••••• o. 5 10 14 Augite ...................... 20 II Hornblende . . . . . . . . . . . . . . . . . 15 Biotite ........ " ........... !J 13 1 1 Chlorite ................ , ... 3 8 19 Sericite n ................ , ... 1 7 3 Olivine ..................... 2 Magnetite ..... , ............. 6b 3 5 Apatite . ,., ....... , ......... 1 I 1 • Includes small amount of hematite and unidentified alteration products. b Largely formed from olivine. l. Olivine trachybasalt, chilled border phase. 2. Granophyre gabbro, average of four thin sections. 3. Granite from center, average of three thin sections. 4. Granite from margin, average of two thin sections. 5. Granite dike cutting granophyre gabbro. II 5 R 16 2 1 134 FHECAMBRIAN GEOLOGY AND GEOCHRONOLOGY gabbro apparently represent magma which worked out from the central intrusion along fractures. The granite is fine- to coarse-grained and dull reel, and because quartz is rare and inconspicuous, the rock resembles a red syenite, similar in appearance to some of the granophyre or "red rock" of thc Duluth complex. Modes of the granite from various occurrences are similar (Table 27). Quartz ranges from 8 to 15 per cent. Alteration has been intensive, anel chlorite and sericite are abundant. Finely divided hematite is disseminated throughout the K-feldspar, and in this respect the granite rcsembles the granophyre of the North Shore region. Accessory minerals are magnetite, apatite, and zircon of the normal type. On the basis of the zircons, Lund (1956, p. 1490) suggested a Keweenawan age for the small intrusives in the Franklin area. The Cedar :Mountain complex represents composite intrusions of mag- mas which were differentiated at depth. The banding of the gabbro is ex- plained by movements of the crystallizing magma at the time of its em- placement. The relatively large amount of granite in the core precludes the possibility of its formation by differentiation after emplacement in the present site. AGE AND ORIGIN Ages for the rocks in the Morton-Sacred Heart region are given in Table 28, Sections A through E. The K-A ages for the Morton gneiss range from 2.45 to 2.57 b.y. Some of this variation must be attributed to analytical error, but the average of .2.5 b.y. is considered to be significant and places the time of formation of the Morton gneiss in the Algoman orogeny. The more or less uniform Sacred Heart granite, which from field relationships is younger than the :Morton gneiss, is a postkinematic intrusion. The aver- age of three K-A age determinations for the Sacred Heart granite is 2.44 b.y., and the Rb-Sr determinations on two of the samples agTee closely. The Rb-Sr ages on two samples of the :Morton gneiss, however, give a similar age of 2.4 b.y., and the K-A age on the Seaforth gneiss is also 2.4 b.y. The frequency of the 2.4 b.y. value in the region shows that this date marks a specific event and that the intrusion of relatively large amounts of Sacred Heart magma into the Morton gneiss at considerable depth re- sultcd in recrystallization of the biotite. Apparently this recrystallization affected the Sr"/Rbs, ratio more than it did the A4°/K"" ratio. It will be noted that the samples of the Morton gneiss (KA-I71 and KA-208) taken near the contact of the Sacred Heart granite (Fig. 29) give ages of 2.47 b.y. This is further support for the hypothesis that the biotite in the Morton gneiss had been affected by the intrusion of the Sacred Heart magma at 2.4 b.y. Subsequent to the main period of Algoman orogeny, the area was not stable, and there were local adjustments. lVlovements resulted in granula- tion and in recrystallization of the granite (KA-24) from the Larsen quarry which closely resembles the Sacred Heart granite to the north. Biotite SOUTHWESTERN MINNESOTA 136 TABLE ~8. AGES FOR ROCKS FROM SOUTHWESTERN lVlINNESOTA K-A Age Rb-Sr Age K,O K 1c) Aj~J KANo. Description b.y. b.y. pct. ppm. ppm. A '"/K''' MORTON-SACRED HEART REGION (A) Morton Quartz Monzonite Gneiss 51B Ridgely township ... 2.49 7.5~ 7.54 2.31 0.306 51F Ridgely township ...... . 12.92 12.95 2.72 .210 14B Morlon quarry (KA-14-l). 2,/>4 8.'20 8.22 2.59 .315 ~.62 .319 2.62 .319 14F Morlon quarry (KA-14-1). 9.24 9.26 2.03 .219 2.02 .218 14B Morlon quarry (KA-H-2). 2J;3 'U3 8.29 8.31 2.62 .315 15B Morton quarry (KA-15-1). 2.+;') G.52 G.53 1.94 .298 151" Morton quarry (KA-1.5-l). 7.87 7.89 1.91 .~W2 1.83 .232 15B Morton quarry (KA-15-2). '2.;);') 7Al 7.42 2.38 .320 107B Northwest of Morton .... 2.57 '2.42 7.99 8.01 2.60 .325 186B North Redwood ......... 2 .• ) 1 8.45 8.47 2.63 .310 171B South of Sacred Heart '2.47 7.22 7.23 2.18 .302 'l08B South of Sacred Heart 'l.47 ,j .84 5.H5 1.76 .301 1.77 .302 A"crage . ......... 2.·50 (B) Ce.dar Moulltai" GI'fl>lophyre Gabbro 195B Cedar lVlountain 1.7;) 4.62 -Ul3 ,'iSS .170 (C) Seaforth Gneiss 210B Seaforth '.uo 8.64 8.66 2.53 .29'2 2A5 .283 (D) Fort Ridgely Granite ,;2B -Forl Ridgely ., ... 2.:30 3.88 3.89 1.03 .2GG ( E) Sacred [-/eart Granite 9B South of Sacred Hearl 2.+2 'lAO 6.53 6.5-1 1.90 .291 91" South of Sacred Heart 8.37 8.39 1.30 .155 13B Northwest of KA-9 2.-l:l '2.45 6.80 6.81 1.98 .291 2.03 .299 23B South of KA-13 'lAG 4.82 4_83 1.45 .300 A\'erage .... '2.44 24E Larsen quarry 'i.30 7.5'2 7 .. 54 1.99 .263 2.01 .267 2.0G .273 1.99 .264 241" Larsen quarry 13.38 13A 1 2.24 .IG7 GRANITE FALLS-lVloNTEVIDEO REGION (F) J\fonte1'ideo Granite Glleiss 27B Montevideo (KA-27-1) . 1.84 5.96 5.97 1.11 0.186 1.09 .183 27F Montevideo (KA-27-I) . 10.46 10.48 1.2-:1 .119 1.22 .116 27B Montevideo (KA-27-2) I.R" 1.74 7 .73 7.75 1.-14 .IS(j ,HB Quarry. Granite Falls (KA-54-1) ........ 1.7;; 6.67 6.68 1.11 .167 1.16 .173 136 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY TABLE 28 - Continued K-A Age Rb-Sr Age K.O K'o A'" KANo. Description b.y. b.y. pct. ppm. ppm. A"/K" 54F Quarry, Granite Falls (KA-54-I) . . . . . . . . . . . . 12.89 12.92 2.04 .158 1.98 .153 1.92 .149 54B Quarry, Granite Falls (KA-M-2) ........... 1.75 l.69 6.31 6.32 l.08 .171 25B North of Granite Falls .. 1.83 5.97 5.98 1.11 .185 1.09 .182 20!lB Southeast of Granite Falls 1.71 7.78 7.80 1.29 .166 1.28 .164 Average .............. 1.79 (G) Garnetiferous Quartz Diorite Gneiss 22B Grani te Falls ........... 1.78 7.12 7.13 1.22 .171 1.29 .180 (H) Granite of Section 28 28B North of Granite Falls ... 1.69 7.32 7.33 1.15 0.156 1.22 .167 29B North of KA-28 ......... 1.69 l.65 7.76 7.78 1.21 .155 1.30 .168 ODESSA-ORTONVILLE REGION (I) Ortonville Granite 55B Bellingham quarry 1.67 8.48 8.50 l.30 .153 1.39 .164 55F Bellingham quarry 14.28 14.31 1.83 .127 56B Bellingham quarry 1.67 9.15 9.17 1.46 .159 108B DlIique Granite quarry .. 1.67 8.44 8.46 1.36 .160 1.33 .157 10gB Odessa ................. 1.79 8.79 8.81 1.59 .180 1.54 .175 57B Milbank granite ......... 1.97 7.62 7.64 1.60 .209 1.54 .202 57F Milbank granite ......... 11.80 11.82 1.92 .162 (J) Quartz-Pyroxene Granulite HB South of Odessa ......... 2.18 5.33 5.34 1.30 .243 (K) Sioux Formation 50 Argillite, Pipestone ..... . 1.20 5.17 5.18 .503 .097 .518 .100 from the granite of the Larsen quarry gives a somewhat lower age of 2.3 b.y. Four argon extractions were made on KA-24 by two different oper- ators. The first two runs (HB) were made in August and in September of 1956. A third run (HWK) was made in December of 1958 and the fourth run (HWK) was made in May of 1959. The minimum and maximum ages from these four runs are 2.28 and 2.33 b.y., and the deviation from the average age of 2.3 b.y. is less than 2 per cent. Although the Fort Ridgely granite gives a similar age of 2.3 b.y., the SOUTHWESTERN MINNESOTA 137 data are inadequate to define an event in this area. In the Franklin area, however, the Morton quartz monzonite gneiss was intruded by basalt dikes and by the Cedar Mountain complex at approximately 1.8 b.y. Be- cause of the similarity in the mineral composition of these basalt dikes to those of the Lake Superior region, and the similarity of the gabbro-gran- ophyre association at Cedar Mountain to the Duluth complex, these rocks had been assumed to be of Middle Keweenawan age. The date of 1.8 b.y., however, for biotite from the granophyre gabbro of the Cedar Mountain complex places it in the Penokean orogeny which is late Huronian or early Keweenawan. It is clear that the Cedar Mountain complex is not con- temporaneous with the Duluth complex which was formed at a much later time, approximately 1.1 b.y. By analogy with the Giants Range and Vermilion granites in north- ern Minnesota, the geologic history in southwestern Minnesota in Early Precambrian time is interpreted to have involved the accumulation of a geosynclinal sequence of beds followed by the intrusion of gabbroic rocks. Near the close of Early Precambrian time (:i!.6-:i!.5 b.y.), the geosyncline is postulated to have foundered, and depression of the sedimentary se- quence to great depths resulted in local melting and generation of a gra- nitic magma which rose during the period of folding, producing the Morton quartz monzonite gneiss or migmatite as a synkinematic intrusion. The xenoliths in the Morton gneiss were derived from gabbroic rocks which were fractured and incorporated in the granitic magma at depth. Subsequent to the main period of folding and formation of the Morton gneiss (:i!.6-:i!.5 b.y.), the Sacred Heart magma was generated and invaded the area under conditions which appear to have involved little stress. The postkinematic magma crystallized to a more uniform granite. The Sacred Heart granite is dated at :i!.4 b.y., which is essentially the time of intrusion of the postkinematic granites in the Rainy Lake region of Ontario. It ap- pears likely that the Fort Ridgely granite was intruded at the same time, but, like the granite which is now exposed in the Larsen quarry, it was subjected to later deformation which reduced the age of the biotite to :i!.3 b.y. GRANITE FALLS-MONTEVIDEO REGION GENERAL STATEMENT The relatively good outcrops in the vicinity of Granite Falls (PI. 5) are among the most instructive in the valley. The basic complex is represented by two units, gabbroic and dioritic gneiss, and a garnetiferous quartz diorite gneiss. These rocks form large outcrops south of Granite Falls. The older gneisses are intruded by the Montevideo granite gneiss, and rela- tively large outcrops of the gneiss occur west of Granite Falls. Although there are many inclusions of diorite-gabbro gneiss in the Montevideo granite gneiss, there are no inclusions of the garnetiferous quartz diorite gneiss. 138 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY Basalt dikes are numerous in this part of the valley and cut the Monte- video granite gneiss as well as the older rocks. A younger granite intru¢les the :Montevideo gneiss and basalt northwest of Granite Falls. Thus the sequence of geologic events indicated by the rock units includes intrusion of the basic complex by the Montevideo granite gneiss, the basalt dikes, and finally by the granite. GARNETIFEROUS QUARTZ DIORITE GNEISS The garnetiferous quartz diorite gneiss forms large outcrops south of the Minnesota River, just south of Granite Falls. The dark-gray rock is distinctly banded and is characterized by abundant pink to red garnet. It contains numerous quartz veins and granitic stringers which, for the most part, follow but also cut across the foliation. The structure appears to be uniform, and the foliation strikes east-west or a few degrees north of east and dips, on the average, 45° S. Only in one place (NW 74 Sec. 4: 39-115) is the garnetiferous gneiss intruded by the granite gneiss. Three small outcrops of the garnetiferous gneiss occur approximately 4 miles northwest of Granite Falls in the SW 74 Sec. 13:40-116, beyond the map limit (PI. 5). In these outcrops the foliation strikes east-west and dips 65° N. A small outcrop also occurs approximately 4 miles southeast of Montevideo, and it appears that the unit may underlie a large part of the valley which is now covered by alluvium. Outcrops of the garnetiferous quartz diorite gneiss are usually weath- ered; however, fresh material is available in a roadcut along State High- way 67 and on the dump of an old prospect pit. The composition of a sample from the pit (Lund, 1956, Table 3, No. 20) is plagioclase (An40), 52; pyroxene, 10; quartz, 16; garnet, 14; magnetite, 5; and accessory horn- blende, biotite, and apatite, 3 per cent. The gneiss in the roadcut is vari- able, and the garnetiferous rock is interlayered with augite-hypersthene diorite gneiss. Some of the layers in the roadcut are exceptionally rich in biotite and contain 10-15 per cent of this mineral (KA-22). GABBRO AND DIORITE GNEISS Inclusions of gabbroic and of dioritic gneiss are abundant in the Monte- video granite gneiss. Commonly the inclusions are layers that conform to the structure of the granite gneiss, and the outcrop pattern of the in- clusions is instructive in detailing the structure of the area. A wide band of gabbro gneiss south of Granite Falls intervenes between the outcrops of Montevideo granite gneiss to the northwest and of the garnetiferous quartz diorite gneiss to the southeast. The structure of the gabbro gneiss con- forms to that of the garnetiferous gneiss, and although the contact be- tween the two units is nowhere exposed, the dip of the gabbro gneiss would bring this unit beneath the garnetiferous gneiss. The gabbro gneiss is gray, medium-grained, and usually distinctly banded. It is variable in texture, structure, and composition. The larger SOUTHWESTERN MINNESOTA 139 masses of the gneiss, south of Granite Falls, are gabbroic, whereas the small inclusions in the Montevideo granite gneiss are dioritic. It seems likely that there has been reaction between the inclusions and the magma similar to that described for the Morton gneiss. The modal analyses (Lund. 1956, Table 1, No. 1,~) for samples from the larger outcrops average calcic plagioclase (An6s), 73; pyroxene, 17; hornblende, 8; and accessory min- erals, ~ per cent. Six samples from smaller masses, for the most part inclu- sions in the Montevideo granite gneiss, show a range of ~~ to 45 per cent plagioclase (An43-G5); 0 to 35 per cent pyroxene; 15 to 76 per cent horn- blende; and 1 to 6 per cent accessory minerals. MONTEVIDEO GRANITE GNEISS Description. The type locality of the Montevideo granite gneiss, 1.5 miles southeast of Montevideo, is readily accessible from U.S. Highway ~1~. The gneiss is pink to red, medium-grained, and characterized by uniform, straight banding. Dark-colored minerals generally make up less than 5 per cent of the rock, except locally where mafic inclusions have been incorpo- rated. Contorted banding, such as characterizes the Morton gneiss, is rarely developed, nor is the variation in grain size so pronounced; there- fore, the Montevideo gneiss is readily distinguished in the field from the Morton gneiss. At the type locality, the foliation strikes about N. 80° E. and dips ap- proximately 60° to the south. Long, narrow bands of gabbro gneiss rep- resent large inclusions in the Montevideo gneiss. A smaller outcrop of the gneiss just to the southeast contains a basalt dike or sill. The most exten- sive outcrops of the Montevideo granite gneiss are in the vicinity of Gran- ite Falls, approximately 1~ miles southeast of Montevideo. The gneiss of the two areas is similar. Composition. Three modal analyses (Lund, 1956, Table 7) of the Monte- video granite gneiss from the Granite Falls vicinity are similar, with micro- cline ranging from 3~ to 40 per cent; plagioclase (Anls), ~4 to 30 per cent; and quartz, 31 to 36 per cent. The mode of the gneiss from the type locality at Montevideo shows a larger percentage of microcline, 61 per cent, and smaller amounts of plagioclase (Anls), 9 per cent, and of quartz, 27 per cent. The average of the 4 modal analyses is given in Table 24. Structure. Lund's detailed mapping of the Granite Falls area shows that the rocks have been folded, and an anticlinal structure, plunging to the east, is indicated (PI. 5). It is suggested that the Montevideo granite magma was intruded into the gabbro gneiss at or just below the contact of the gabbro gneiss with the overlying massive unit of garnetiferous quartz diorite gneiss. South of the Minnesota River, the garnetiferous quartz diorite gneiss strikes slightly north of east and dips at angles of 40 to 65° to the south, whereas the small outcrops, previously mentioned northwest of Granite Falls, strike east-west and dip 65° to the north. The latter out- crops represent the northern limb of the anticline. 140 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY A second fold is indicated in Sections 10 and 11, 1.5 miles south of Granite Falls, where a band of the Montevideo gneiss with included gab- bro gneiss, south of the Minnesota River, dips at high angles, 50-85°, to the north. North of the river, the foliation of the garnetiferous quartz diorite gneiss dips at angles of 35-50° south. The Minnesota River thus flows along the axis of the syncline. BASALT DIKES Numerous basalt dikes cut the various gneisses of the Granite Falls- Montevideo area (PI. 5). Although some appear to be concordant with the structure of the rock, and therefore might appropriately be called sills, they commonly show some discordance, as can be seen on the geologic map. For this reason all are included in the classification of dikes. Most of the dikes are basaltic, but they have not been studied in detail, and some may be lamprophyres. The dikes range from less than one inch to many feet in width. The narrow dikes are aphanitic, but the larger ones commonly grade from fine-grained chilled contacts to coarse diabasic cen- ters. The principal minerals are labradorite, augite, magnetite, and inter- stitial quartz-K-feldspar intergrowths. Late stage or hydrothermal altera- tion is responsible for the development of hornblende, epidote, sericite, calcite, and other secondary minerals. LATE GRANITE OF SECTION 28 Small granite intrusives represent the youngest known igneous activity in the region. The relative age relations are well established in the field and are easily seen in the railroad cuts north of Granite Falls (PI. 5). Lund (1956, p. 1489) notes a small outcrop of granite along U.S. Highway 212 northwest of Granite Falls in Sec. 13: 40-116. A third occurrence, southeast of Granite Falls in Section 11 just wuth of the pond (PI. 5), are dikelets of granite and of pegmatite. These are a few inches wide and cannot be mapped. They cut a basalt dike and the Montevideo granite gneiss. Un- doubtedly there are additional occurrences of this younger granite, but in the field the granite can be readily identified only when it is found cutting the basalt dikes. Small granitic stringers and dikes in the gabbro gneiss and in the garnetiferous quartz diorite gneiss may be of the same age as the granite cutting the dikes. The granite of Section 28 is well exposed in the railroad cut along the right-of-way of the Chicago, Milwaukee, St. Paul and Pacific Railroad. The granite appears to have been localized by an earlier intrusion of basalt in the :lVIontevideo granite gneiss. The Montevideo granite gneiss, well exposed a few feet south in a cut along the Great Northern Railroad tracks, contains inclusions of the gabbro gneiss so that representatives of the prin- cipal rock types can all be seen within a small area. The younger granite is gray to pink and medium-grained. The basalt was brecciated and incorporated in the granite, forming an unusual ag- SOUTHWESTERN MINNESOTA 141 FIGURE 31. - Igneous agglomerate formed by intrusion of granite into basalt . Rai lroad cut in Section 28, north of Granite Falls. FIGURE 32. - Close-up view of the agglomerate shown in Figure 31. Note the rounded pieces of basalt in coarse-grained granite. Railroad cut in Section 28, north of Granite Falls. Determined age is 1.7 b .y. 142 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY glomerate. The fragments have been somewhat rounded, probably by attrition as well as by reaction with the magma, and the rock resembles a boulder conglomerate (Figs. 31, 32). Quartz is abundant, 32 to 34 per cent, and feldspar makes up 50 to 60 per cent of the rock, but the proportions of micro cline and oligoclase vary greatly in different samples. Biotite, hornblende, and magnetite are the principal dark-colored minerals. In an outcrop in a pasture just north of the county road and north of the rail- road cut, medium-grained pink granite is relatively free of basalt inclu- sions. Lund's mode (1956, p. 1486, No. 37) of this rock gives quartz, 32; K-feldspar, 30; plagioclase (An20), 26; biotite, 11; and accessory magnetite, apatite, zircon, and sphene, 1 per cent. AGE AND ORIGIN The radioactivity ages for rocks of the Granite Falls-Montevideo re- gion are given in Table 28, Sections F, G, and H. The K-A ages fall within the limits of 1.69 b.y., for the granite of Section 28, and 1.85 b.y. for the Montevideo granite gneiss at the type locality near Montevideo. The average age for the Montevideo granite gneiss is 1.8 b.y., giving a differ- ence of approximately 100 million years between the time of folding and the time of intrusion of the late granite. The three Rb-Sr age determina- tions, ranging from l.65 to l.74 b.y. lend support to the time difference indicated by the K-A ages. The ages are approximately those found for the intrusives in east-central Minnesota and mark igneous activity which accompanied the Penokean orogeny at the close of Middle Precambrian time. The interpretation of the ages, in terms of the geologic relationships that are known from field and laboratory studies, is somewhat uncertain, if not speculative. The oldest rocks in the area were assigned by Lund to the basic complex; however, in the earlier work it was thought that the Montevideo granite gneiss and the Morton gneiss were of the same age, and the gneisses were assigned to the Minnesota Valley granite series. The difference in ages found for the Montevideo gneiss and for the Morton gneiss suggests the possibility that the rocks of the basic complex may not be of the same age in the two areas. The oldest rock would appear to be the garnetiferous quartz diorite gneiss. The large folds in the Granite Falls area suggest the possibility that these rocks represent an original sedimentary series, but little evidence was found to substantiate this hypothesis. Some of the observations which conflict with this interpretation may be briefly considered. The garnetifer- ous quartz diorite gneiss south of Granite Falls is a metamorphic rock (garnet-amphibolite facies). If the assumption is made that an original sequence of sedimentary rocks was metamorphosed at considerable depth to produce the Montevideo granite gneiss and the garnetiferous quartz diorite gneiss at the same time, it is hard to understand why garnet was not developed also in the dioritic and gabbroic gneisses which are similar SOUTHWESTERN MINNESOTA 143 in composition to the garnetiferous quartz diorite gneiss and are found as inclusions in the Montevideo granite gneiss. Lund found mechanical effects of shearing and granulation in the Montc- video granite gneiss, and there is a small amount of garnet in the rock in outcrops along County Road 4, east of U.S. Highway 212, southeast of Montevideo. These features suggest later deformation, but the garnet ap- pears to be a local development. The Montevideo granite gneiss is fairly uniform in composition over a distance of 15 miles, and there is a marked absence of textural features which might be interpreted to indicate a pos- sible sedimentary origin. The uniformity in composition of the leucocratic gneiss and the reactions which appear to have taken place, as is indicated by the differences in composition of inclusions related to their size, argue that the Montevideo granite gneiss is of magmatic origin. The metamorphism which formed the garnetiferous quartz diorite gneiss is attributed to an earlier period, and the age of 1.8 b.y. obtained for sam- ple KA-22 is not related to this period of metamorphism but to the time of the intrusion of the Montevideo granite magma. It is postulated that previous to this there was an intrusion of gabbro beneath or into the gar- netiferous quartz diorite gneiss, forming a sill-like mass. The Montevideo granite magma, derived by melting of sediments at depth, was intruded into the gabbro at or near the present base of the garnetiferous quartz diorite gneiss. At the time of the granite intrusion or shortly following, 1.9-1.8 b.y., the area was folded. It was at this time that the biotite in the garnetiferous gneiss was recrystallized, and thus it gives an age similar to that of the Montevideo granite gneiss. The period of folding was followed by intrusion of basalt dikes. Two periods are indicated by cross-cutting dikes mapped by Lund. Northwest of Granite Falls in Sees. 19 and 24 (Lund, 1956, PI. 3), an east-west dike is cut by later dikes which trend northeast. West and northwest of Granite Falls the northeast trend of the basalt dikes is conspicuous and suggests some type of tensional relief following the folding. In Section 28, basalt was brecciated and intruded by granite. The Montevideo granite gneiss (KA-25) from just south of the basalt intrusion is dated at 1.8 b.y., whereas the granite (KA-28, 29) is dated at 1.7 b.y. The basalt dikes, which are post-Montevideo granite gneiss and pre-granite, were intruded between 1.8 and 1.7 b.y. Thus it seems reasonable that their time of intrusion was about the same as that of the Cedar Mountain complex in the lVlorton gneiss south of Franklin which is dated at 1.75 b.y. In summary, the radioactivity ages for the Granite Falis-lYlontevideo region indicate folding and intrusion at approximately 1.8 b.y., approxi- mately the time of the early phase of the Penokean orogeny in east-central Minnesota. The Montevideo granite gneiss is a synkinematic intrusion. Pre-existing rocks were metamorphic rocks, such as the garnetiferous quartz diorite gneiss of the Granite Falls area, and dioritic or gabbroic rocks intrusive into this series. In the late stage of folding, late kinematic 144 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY or postkinematic basalt dikes were intruded, followed by intrusion of granitic magma. This sequence of events probably took place within a time interval of the order of 50 to 100 million years. ODESSA-ORTONVILLE REGION GENERAL STATEMENT Lund used the name Ortonville granite for the outcrops between Odessa and Ortonville but noted that a similar rock type occurs in a large area between Ortonville and Milbank, South Dakota. The granite of this re- gion is variable in texture, color, and structure, and Lund distinguished two varieties but did not use specific names. In the area of main outcrop between Ortonville and Odessa, the granite ranges from pink to deep ma- roon red, and from medium-grained to coarse pegmatitic. Porphyritic texture is developed locally. Streaks and segregations of quartz-feldspar pegmatite make up a large part of the rock and give it a banded or mar- bled appearance. Dark inclusions are plentiful in some localities, and the rock commonly shows a pronounced gneissic structure. Polished surfaces of the gneissic granite are striking in the contrast between the medium- grained, dark-red rock and the lighter-colored, large, augen-like pegmatitic areas of coarse feldspar and blue quartz. The granite in the many small knobs in the valley southeast of Odessa and north of Bellingham resembles the granite between Odessa and Orton- ville, but mafic inclusions appear to be more numerous, and the rock is more porphyritic and less commonly foliated or banded. In the present report the term Ortonville is used in the sense of Lund's classification for all the granites in the upper part of the Minnesota Valley. The granite be- tween Odessa and Ortonville is here referred to as the Odessa type, whereas the granite southeast of Odessa is called the Bellingham type. BASIC COMPLEX Larger inclusions are rare in the granite between Odessa and Ortonville; however, southeast of Odessa there are numerous inclusions as well as small isolated areas which are sufficiently large to be shown on a geologic map (Fig. 33). The inclusions in the Bellingham granite are fine-grained, sugary, schistose, and banded rock. Outcrops in the northeast corner of Sec. 9: 120-45 are layered and are a metamorphosed sedimentary sequence. The layers strike N. 7° E. and are essentially vertical. A similar dark- colored, banded rock crops out on the west side of U.S. Highway 75 in the southeast corner of Section 4 (Fig. 33). This rock was described by W. S. Bayley (Diller, 1898, p. 358), who classified it as a quartz norite gneiss. A modal analysis by Lund (1956, p. 1481, No. 19) gives plagioclase (Ans7), 57; pyroxene, including both augite and hypersthene, 15; biotite, 5; quartz, 21; and accessory magnetite and apatite, 2 per cent. Lund describes com- plicated folding and siliceous stringers which are deformed into small ptygmatic folds in the mafic inclusions. --,- ------- ------ -------- ! EXPLANATION QUATERNARY ~ Surficial deposIts (Pleistocene dnll and Recent alluVium) PRECAMBRIAN 5=:l SOUTHWESTERN MINNESOTA , Ouarry , Strike and dip of foliation 145 Basalt dike Somp'~d IOCalify ! With A,~,odote I In billion years t--' -+------+-....".,,.,,;;,-..:; ~_¢jI!V__'~,___; Ortonville granite (Red, coorse /0 porphYfllic, ..,;/h obvndonl inc/uslons of Quartz dIorite gnelss(qdg) mopped separately where scale permlls) SCALE U'IU 1.7 og ... 17 FIGURE 33 . -Map of the Minnesota River Valley southeast of Odessa showing location of dated samples. ODESSA GRANITE Osd The granite between Odessa and Ortonville is commonly red, but th ere are pinkish-gray to reddish-brown phases, and some are deep maroon red . It is generally medium- to coarse-grained or pegmatitic. Coarse-grained porphyritic granite of a dark-red color is quarried one mile south of Odessa. Inclusions are relatively few, and the rock shows no pronounced structure except for an alignment of large K-feldspar crystals. An east- west basalt dike cuts the granite. In the northwestern part of this quarry, stripping has exposed Cretaceous silts which lie unconformably on the granite. Large, well-rounded boulders of transported granite and other rock types mark the unconformity. Fossils from the Cretaceous sediments, collected by Robert C. Bright, include Inoceramus sp., unidentified pele- 146 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY c,Vpods, Ptycodont shark teeth, shark vertebrae, fish teeth and scales, reptilian teeth, and plant fragments. Pleistocene deposits overlying the Cretaceous sediments and the granite are mainly silt and clay beds con- taining mollusks and plant fragments. The average of a modal analysis of the granite from this quarry and of a similar granite from a quarry 1 mile west of Odessa is given in Table 24. A recent chemical analysis (Table 2.5) of the Ortonville granite, obtained through the courtesy of the Cold Spring Granite Company, represents the rock from a quarry in Sec. 26: 121-46. The analysis is similar to one of the l\10rton quartz monzonite gneiss from the main quarry at Morton. The structure of the granite between Odessa and Ortonville is variable. The foliation or alignment of phenocrysts is commonly east-west, but variations are numerous, and locally the strike may be northwest or north- east. BELLINGHAM GRANITE There are a number of small quarries south of Odessa and north of Bel- lingham (Fig. 33). Although the granite resembles the Odessa type, mafic inclusions appear to be more numerous, and the rock is commonly porphy- ritic. The color ranges from red to dark-reddish brown. The average of two modal analyses of the granite from a quarry in the SW 1/4 Sec. 15: 120- 45 is given in Table 24, and a chemical analysis of the granite from a small quarry V2 mile west, in the SE cor. Sec. 16, is given in Table 25. MILBANK GRANITE Coarse reddish-brown granite is exposed in a quarry 1 mile south of U.S. Highway 12, approximately 7 miles southwest of Ortonville in Grant County, South Dakota. There are a large number of inclusions in the granite, and skialiths indicate reaction with magma. Some of the inclu- sions in the granite resemble stratified rock and are suggestive of sedi- mentary rather than igneous parentage. The granite in the l\1ilbank area is similar to that between Ortonville and Odessa. AGE AND ORIGIN The K-A ages determined for samples from the Odessa-Ortonville re- gion (Table 2S) show a wide range, from l.7 to 2.2 b.y. Biotite from the quartz-pyroxene granulite of the basic complex gives the oldest age of 2.2 b.y., and it seems quite likely that this is a minimum value and that the rock probably originated through high-grade metamorphism of a sedimentary sequence during folding in the Algoman or an earlier orogeny. The granite from the quarry 1 mile south of Odessa was dated at l.S b.y.: however, the Milbank granite is dated at 2.0 b.y. Biotite concentrates from the granites of the Bellingham type (samples KA-55, 56, and lOS) give an age of l.67 b.y. The Odessa-Ortonville region is a disturbed area, with a number of SOUTHWESTERN MINNESOTA 147 periods of metamorphisms and granite intrusions, making the K-A ages difficult to interpret. The Bellingham granite may be of the same age as the late granite of Section 28, north of Granite Fall s. SIOUX FORMATION DESCRIPTION The Sioux formation (White, 1870) underlies a large area in Minnesota, South Dakota, and Iowa. Outcrops are found in an east-west belt 200 miles long, extending from New Ulm, Minnesota, to Mitchell, South Dakota, and about 50 miles wide, from Flandreau and Canton, South Dakota (Fig. 34). The formation is composed chiefly of red, fin e-grained, hard quartzite, which is a prominent ridge maker, and the natural exposures are usually cuestas. Gray to light-pink quartzite is quarried at Jasper in Rock County. The Sioux formation also contains poorly cemented sandstone, con- glomerate, and beds of silty and very fine-grained shale with little or no quartz. An argillite layer, 15-18 inches thick (Fig. 35), in the quartzite just north of Pipestone, was used by the Indians in the manufacture of pipes, and this is the origin of the name Pipestone both for the rock and for the county and city. George Catlin visited and collected pipestone from the Indian quarry in 1836, and the pipestone is also called catlinite (Jackson, 1839). There are a number of easily accessible outcrops of the Sioux formation in the vicinity of New Ulm in Nicollet County. The conglomerate at the base of the Sioux formation forms a ridge approximately 1000 feet long southeast of New Ulm, in Sec. 27:110-30. The beds strike N. 20 0 E. and dip 15-20 0 SE. P ebbles and cobbles are largely quartz, jasper, and chert. SOUTH DAKOTA I Ortonville I WIS CONS IN I I ~. I KA-50 ~ II 1.2 b .y. ~ Ridgely New Ulm KA-192 Red Wing 435 m.y. KA-1I3, KA_114'o 450,430m .y. 435 m.y. Flandreau. '\PiPestone I. Jasper • Sparta ° Milchell • _,Garretson MINNE SOTA Sioux Foils. Cor~~ _____________ _ o 50 E3 E3 E3 IOWA 100 150 MILES . Readstown 410m.y. FIGURE 34. - Map showing location or dated sample (KA-50) or pipestone rrom the Sioux formation and of Cambrian glauconite samples. KA-1l3 and KA-1l4 are from the Franconia formation ; KA-192 is rrom the Jordan formation . The ages in Wisconsin are ror glauconite from the Franconia formation ; at Sparta. K-A age by Wasserburg et al. (1956), at Readstown , Rb-Sr age by Herzog et al. (1958 ). 148 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY FIG URE 35 . - Bed of pipestone in quartzite of the Sioux formation. Pipestone Quarries National Monument. The pipestone gives a K-A age of 1.2 b .y . which may be the time of regional folding. Two hundred yards west of the ridge are some small outcrops of weathered, coarse, porphyritic granite. Southeast of this locality, in the vicinity of Courtland, the quartzite is quarried for crushed rock. Baldwin (1949) estimated the thickness of the Sioux formation in expo- sures along Split Rock Creek near Garretson, South Dakota, to be approxi- mately 3000 feet. This is the order of magnitude of an earlier estimate for this section by Irving (1885, p. 201). Todd (1895, p. 35) thought that Ir- ving's estimate was in excess and suggested a maximum thickness of 1500 feet for the Sioux formation . The maximum known thickness in Minnesota i approximately 500 feet (Thiel, 1944). The Sioux formation is folded, but the regional structure is poorly known. In the New Dim-Courtland area, the strike of the beds ranges from N. 70° E. to N. 70° W., and the dip is also variable and may be as high as 30°. Similar variable dips are found in western Minnesota and in South Dakota. PIPESTONE OR CATLINITE The most detailed description of pipestone or catlinite is given by Berg (1938), who studied the Sioux quartzite outcrop north of Pipestone in the area that is now aN ational Monument. Here the beds dip east at approxi- mately 15 0 . The pipestone is a blood-red argillite, in part replaced by a white to yellowish-white mineral which Berg identified as pyrophyllite. SOUTHWESTERN MINNESOTA 149 The pyrophyllite has replaced the top and bottom of the catlinite bed to a depth of approximately one inch, and also occurs along the partings parallel to the bedding. The principal mineral is sericite, and in addition to the pyrophyllite, Berg identified diaspore, hematite, specularite, and pyrite. He concluded that the pipestone represents a fine-grained argil- laceous layer, "converted into secondary mica (sericite) during consolida- tion and metamorphism." The sericite was partly altered to pyrophyllite which in turn was altered to diaspore. A thin, 6-10 inch, layer of fine- grained, red argillite occurs in the quartzite in Sec. 35: 110-30, southeast of New Dim. The layer resembles the pipestone in the quartzite north of Pipestone but is composed of sericite and very fine-grained quartz. AGE AND CORRELATION The pipestone from the Sioux formation (KA-50) has been dated by the K-A method at 1.2 b.y. (Goldich, Baadsgaard, Edwards, and Weaver, 1959, p. 660). This sample came from Berg's collection, and material from the same specimen was used in the chemical analysis made by R. B. Elle- stad (Berg, 1958, p. 261). Weaver determined that 60 per cent of the sam- ple is composed of 2M-illite. The sample used in the age determination was crushed and sieved (40-60 mesh) to eliminate fine materials. The average of two determinations of K 20 is 5.17 per cent, which is somewhat less than the 5.62 per cent of K 20 found by Ellestad, and the difference is undoubt- edly due to the method of sample preparation, namely the elimination of the fine material in the samples used for the argon determination. Two extractions of argon were made, and the average gives an age of 1.2 b.y. (Table 28). This is considered to be a minimum age for the Sioux forma- tion and very likely dates the time of folding. Grout and co-authors (1951, p. 1051) report that some of the pebbles in the conglomerate beds of the Sioux formation resemble the granule- textured iron formation of the Biwabik, and they placed the Sioux forma- tion "tentatively between the Animikie and the Keweenawan." The por- phyritic granite beneath the conglomerate southeast of New Dim is badly weathered, and material for dating could not be obtained; however, the granite at Fort Ridgley (Fig. 26), 15 miles northwest of New Dim, is dated at 2.5 b.y. At Corson, South Dakota (Fig. 34), 30 miles south- southwest of Pipestone, the Sioux quartzite is intruded by diabase (Todd, 1895, p. 36; Beyer, 1897, p. 82). Sardeson (1908) reported dikes which cut the Sioux formation in the New Dim area. These dikes may be of Middle Keweenawan age. The Sioux formation merits further study because it seems likely that the formation is of early Keweenawan age and was folded at approximately 1.2 b.y. This date roughly marks the beginning of the Keweenawan igneous activity in the North Shore region. 6. DEVELOPIVIENT OF A PRECAMBRIAN CLASSIFICATION INTRODUCTORY STATEMENT The development of the geologic time scale, familiar to every student of geology, has been a long and difficult task. The success that has followed with the use of fossils for relative age determinations and for correlation probably is the outstanding achievement of geological science. The Pre- cambrian rocks, however, are essentially devoid of a useful fossil record, and in the absence of reliable methods of correlation, a world-wide classi- fication has not been achieved. The development of a classification for the Precambrian rocks of Minnesota roughly parallels that of most regions. Various rock units have been named and renamed; original usages have been modified or completely changed; and efforts to correlate with other regions have won only modest acceptance. A review of Precambrian literature is a major task quite outside the scope of the present report, but there are a number of reviews that are readily available to the interested reader. The older writings pertaining to the Canadian Shield were reviewed by Adams (1915) and Coleman (1915). Bulletin 769 of the United States Geological Survey (Wilmarth, 1925) compares the geologic classification of the Geological Survey with others in use at the time. The original definitions of the terms era, period, and epoch were compiled, and this material is helpful in understanding the evolution of the Precambrian terminology. A detailed review of geo- logic work in Canada from 1897 to 1912 is given by Young (1932, 1933). Professional Paper 184 (Leith, Lund, and Leith, 1935) summarizes the work of the United States Geological Survey and includes a geologic map of the Lake Superior region in the United States. Wilson (1939) has given an excellent summary of the geology of the Canadian Shield and has dis- cussed some of the problems of correlation and nomenclature. Two recent volumes on the Precambrian geology of Canada, each including a number of papers, are The Grenville Problem (Thomson et al., 1956) and The Proterozoic in Canada (Gill et al., 1957). Early efforts in the development of a classification of Precambrian rocks attempted to follow the principles developed for the younger rocks. By analogy with Cenozoic, Mesozoic, and Paleozoic, the terms Proterozoic and Archeozoic were introduced. The distinctions between rock units, time-rock units, and time units usually were followed. The lack of fossils, however, made the Precambrian rocks uninteresting to the stratigraphers, and the study of the older rocks was more or less relegated to the economic geologists and petrologists. It must be conceded that there is much disagreement concerning the 150 DEVELOPMENT OF A PRECAMBRIAN CLASSIFICATION 1.51 usefulness of the terminology that has been evolved, and in view of the complicated structural and metamorphic history of the Precambrian rocks, it is not remarkable that differences of opinion should have arisen about their ages and relationships. It should also be admitted that stratigraphic rules of nomenclature designed for the younger rocks are at times difficult to apply to Precambrian rocks and have not always been followed. Time terms and rock terms have been used interchangeably. For example, al- though the Animikie group of Ontario and Minnesota is a well-defined lithic unit, it has been convenient to speak of the time of deposition or of the seas in which the strata were deposited as Animikie or Animikian. These usages can be found in the literature, and generally no misunder- standing results, but the world-wide correlation of iron formations on the basis of lithologic similarities or on succession of similar beds is a much more dangerous undertaking. Correlations indicate time equivalence, and in the absence of fossils, correlation of Precambrian rocks has been dif- ficult to achieve. TIME SCALE Radioactivity dating affords the best if not the only possibility for cor- relating Precambrian rocks. The early chemical methods of uranium-lead dating could not take into consideration the nonradiogenic lead in the analyzed minerals; nevertheless, these early efforts were far superior to the attempts to measure geologic time on principles such as the rate of accumulation of sediments, the rate of accumulation of salt in the oceans, and similar methods. PRINCIPLES The establishment of a time scale and the establishment of a classifica- tion for the Precambrian rocks are not one and the same problem, but they are closely related. It is possible today to set up a time scale with an arbitrary interval, for example of 500 million years, to give convenient simple numbers at time boundaries of 1.0, 1.5, '2.0 b.y., and so forth. The intervals starting with the youngest could be called Precambrian L '2, 3, 4, etc. or by some other equally simple system. Geologically, such a time scale probably would have little to recommend it. Radioactivity dating as a useful geologic tool requires no defense and little elaboration, but it is recognized that if geologic mapping is reduced to collecting and dating samples, little progress can be envisioned for the near future. Geologic time should be independent of special events, but like ordinary time or human chronology, there are some special considerations. A time scale as well as a classification of Precambrian rocks must be coordinated with geologic features that are recognizable to the field geologists. Orogenies. Periods of mountain building accompanied by metamorphism and igneous activity can be dated and basically become of great impor- tance to Precambrian chronology. Lawson (1913a) used Laurentian and 152 PRECAMBRIAN GEOLOGY AND GEOCHRONOWGY Algoman as time terms for two periods of granitic intrusive activity and inserted them in his stratigraphic column (Table 11). An objection to this practice was earlier raised by Lane (1905), whose position was well taken. The Laurentian and Algoman are periods of orogeny accompanied by metamorphism and igneous activity, and in the present report, the periods of orogeny are shown independently of the time scale (Table 2). The granite of Laurentian age is defined as being post-Keewatin and pre-Seine (Knife Lake). Similarly, the Algoman granite is defined as being post-Seine and pre-Animikie; thus time connotations are given to lithic units. It is recognized, however, that during the intrusion of the Lauren- tian granite, rocks which are lithologically indistinguishable from typical Keewatin rocks may have been deposited. In relationship to such possible rocks, the Laurentian granite is not post-Keewatin. Marsden (1955, p. 113) clearly anticipated the Penokean orogeny of this report in writing "a post-Huronian orogenic belt extending eastward from the Cuyuna Range may be responsible for the deformation and meta- morphism of the Thomson Slates southwest of Duluth, Minnesota." He classified granites, which are assigned to the Penokean, as being of "post- Huronian and pre-Keweenawan age." It is evident that he meant these granites are intruded into rocks which locally are classified as Huronian and are overlain by rocks which are locally assigned to the Keweenawan. Calling these granites Penokean eases some problems of nomenclature. Unconformities. Orogeny leads to erosion and to the development of unconformities which geologically are of great value. The Animikie group is defined and identified on lithologic characters, but everywhere in Min- nesota and in Ontario the Animikie strata rest with angular unconformity on rocks of Early Precambrian age. Lawson was impressed with the im- portance of unconformities and with the time involved in their forma- tion. He gave formal names, Epilaurentian Interval and Eparchean In- terval, to two periods of time during which unconformities were formed. The Eparchean Interval during which the Algoman Mountains were more or less eroded was calculated by Lawson to have a duration of over 200 million years, and a peneplain was thought to have been formed over most of the Canadian Shield. It is obvious that the materials eroded must have been deposited in adjacent areas, but there remains the problem of recognizing this material in the Middle Precambrian sedimentary record. It is not here intended to make a case for world-wide orogenies or world- wide peneplains, and it is clearly recognized that neither periods of moun- tain building nor of peneplanation can be considered to be planes in time; however, because orogenies and unconformities can be recognized and the former dated, they are of primary importance in any scheme of Precam- brian classification. Correlation. Erosion and deposition are complementary, and the rec- ords of these processes cannot be found in a single restricted region. The larger record must be pieced together from work in many regions. In the DEVELOPMENT OF A PRECAMBRIAN CLASSIFICATION 153 discussion that follows, correlations of the events of Precambrian time in Minnesota and adjacent areas in the Lake Superior region with dated events in other parts of North America and on other continents are sug- gested. No attempt has been made to include all the data available in the literature. It is hoped, however, that a sufficient number have been se- lected to give the reader some idea of the progress that is being made. The potassium-argon and rubidium-strontium ages published by other in- vestigators have been adjusted for decay constants so that they are com- parable to the results of the present study. It is emphasized that the un- certainties in decay constants, in analytical procedures, and in geologic interpretations cannot be ignored and that the indicated correlations are tentative. NOMENCLATURE The United States Geological Survey recognizes only local or provincial time divisions within the Precambrian. The Geological Survey of Canada recognizes two subdivisions, the Archean and the Proterozoic. The Min- nesota Geological Survey found that a three-fold division is useful, and this classification has been revised in the present investigation. The eras are referred to as Early, Middle, and Late Precambrian, with time bound- aries of 2.5 and 1.7 b.y. It will be seen in the discussion that follows that the Algoman and Penokean orogenies can be correlated, at least approxi- mately, with similar orogenies in many parts of the world. :Many of these orogenies or cycles of sedimentation and metamorphism have been given local names, and this seems to be a useful practice. 'Vith the refinement of the dating technique, correlation of these periods will be practicable. That further work will show the desirability and the necessity for re- vision of the present classification is beyond question, and in view of this condition no new terms have been introduced in this report, although some of the old names have been used in a restricted sense. EARLY PRECAMBRIAN It may be that a specific name such as Archean is desirable. For exam- ple, it is somewhat awkward, if not confusing, to speak of early Early Pre- cambrian, whereas early Archean is simple and intelligible. In part this difficulty is avoided by the use of period names. Ontarian. A committee of the Royal Society of Canada (Alcock, 1934) suggested Laurentian and Timiskamian as time divisions of the Archean (Table 29). Laurentian, however, is used to designate a period of orogeny and is unsuitable for a system. The Geological Survey of Canada (Har- rison, 1957, Table 6, p. 27) has used Keewatin and Timiskaming as period names; however, Keewatin is used to define a specific lithic unit, and there- fore Ontarian is retained in this report for informal use in the sense of Lawson's (1913a, b) precedent. The Coutchiching and Keewatin groups are included in the Ontarian. 154, PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY TABLE 29. CLASSU"ICATION OF THE PRECAMBRIAN RECOMMENDED BY THE CANADIAN NATIONAL COMl'>IITTEE ON STRATIGRAPHICAL NOMENCLATURE (Alcock,1934,p.118) Precambrian r Keweenawan Proterozoic i Archean l Huronian J Timiskamian 1 Laurentian { Upper Keweenawan Lower Keweenawan ~ Upper Huronian l Lower Huronian { Upper Timiskamian Lower Timiskamian r Upper Laurentian l Lower Laurentian Laurentian Orogeny. The time boundary between the Ontarian and the Timiskamian remains to be delimited, but it is marked geologically by the Laurentian orogeny. In the earlier classification of the Minnesota Geological Survey the Laurentian, following Lawson's concept, was con- sidered a major orogeny, and the Earlier and Medial rocks were separated on this basis (Table 1). The present study, however, indicates that the Laurentian and Algoman orogenic movements were not separated by a billion or more years and favors the interpretation of the Laurentian fold- ing as an early phase of the greater Algoman orogeny. Cooke (1926, p. 32) anticipated this possibility, and concluded that "The pre-Timiskaming folding, important as it was, was only a minor movement compared to the great folding that succeeded the Timiskaming deposition and closed the Archean era." Tin7;iskamian. A sedimentary cycle of the Timiskamian period is repre- sented by the Knife Lake group of Minnesota which is correlated with the Seine group of Ontario. In the vicinity of Steeprock Lake, Ontario, a se- quence of conglomerate, graywacke, argillite, carbonate rocks, iron forma- tion, tuffs, and flows is referred to the Steeprock group. The age of these rocks is controversial. The group lies unconformably on a granitic com- plex that is assigned to the Laurentian and was intruded by granite classed as Algoman. A sample (KA-298) of the younger granite has been dated at 2.56 b.y. (Table 14), giving a minimum age for the Steeprock metasedi- ments. Whether the group should be correlated with the Seine or with older rocks is uncertain. Algoman Orogeny. The lack of positive geologic control and the uncer- tainties of the dating techniques, in regard to analytical errors and decay constants, make it impossible to place close limits on the time of the Algo- man orogeny. The present age data suggest an interval of 200 million years DEVELOPMENT OF A PRECAMBRIAN CLASSIFICATION 15.'; between the development of the orogenic gneisses and accompanying syn- kinematic intrusive rocks and the postorogenic or postkinematic granite plutons. Correlation. The Algoman Mountains extended eastward into Quebec where the orogenic belt is truncated by the younger Grenville orogenic belt (Fig. 1). Aldrich and Wetherill (1960) reported K-A ages of 2.54 b.y. for a pegmatite near Hearst, Ontario; 2.39 b.y. for granite near Timmins; and 2.52 b.y. for granite near Kirkland Lake. Their Rb-Sr ages for these samples are 2.44, 2.36, and 2.32 b.y., respectively. The general region of the Canadian Shield that was affected by the Algoman orogeny commonly is referred to as the Keewatin or Superior province. Early Precambrian rocks, 2.5 b.y. or older, have been reported from many other parts of the world, as for example, from the Bighorn Basin of Wyoming and Montana (Gast et aZ., 1958), from Australia (Jeffery, 1956), from Minas Gerais, Brazil (Hurley, personal communication, 1959), and from Africa. Holmes and Cahen (1957) have summarized the African data available in July 1956. They proposed seven Precambrian cycles of which the Shamvaian, dated at 2.65 b.y. (U-Pb), may be correlative with the general period of Laurentian-Algoman orogeny in the Canadian Shield. Two older unnamed cycles also are suggested, 2.8-3.1 b.y. and 3.2-3.4 b.y. Extensive geochronologic studies are now in progress in the U.S.S.R. Potassium-argon dating of the rocks of the Baltic Shield indicate events older than 3.0 b.y., and Gerling and Polkanov (1958) have suggested a cycle of sedimentation and metamorphism from approximately 2.8 to 3.4 b.y. which they call the Katarchean cycle. MIDDLE PRECAMBRIAN The Middle Precambrian era as used here corresponds to the early part of the Proterozoic of the Canadian Geological Survey. Sedimentary rocks of this age are of limited known occurrence in :lVIinnesota, and subdivision of the 800 million years of Middle Precambrian time is not possible. The Animikie group, formerly assigned to the Later Precambrian, and the Thomson formation are now placed in the Middle Precambrian (Table 2). A possible correlation of the Thomson with the Virginia formation is in- dicated. The iron formations and related rocks of northern Michigan that were called Huronian are now referred to the Animikie series by the United States Geological Survey (James, 1958) and are assigned to the Middle Precambrian; however, the Federal Survey uses the terms Lower, :lVIiddle, and Upper Precambrian as informal divisions of rock units. As James points out, it is inevitable that these terms will be interpreted as time units, and he uses them informally in this sense. In the table of lithologic se- quence of the Precambrian rocks, for example, James (1958, Table 1, p. 30) notes that the granitic rocks which were intruded into the Animikie series have a probable age of at least 1.4 b.y. 156 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY Huronian. It appears probable that the rocks of the type Huronian will prove to be in large part of Middle Precambrian age. Studies of the rocks of the Sudbury region, now being conducted by the geochronology groups of the Carnegie Institution of Washington and of the Massachusetts In- stitute of Technology, show that the region is a disturbed one, and there are some marked differences in the Rb-Sr and K-A ages for the same sam- ples (Fig. 9). Feldspar from the Copper Cliff rhyolite has been dated (Rb- Sr) at 2.2 b.y., and biotite from the same rock has been dated (K-A) at 2.1 b.y. (Aldrich and Wetherill, 1958). The term Huronian was not used by the :Minnesota Geological Survey because of the uncertainty of the correlation with the type locality and because the former assignment of the Knife Lake group to the Lower Huronian and of the Animikie group to Upper Huronian resulted in the separation of the Huronian rocks by the great unconformity at the base of the Animikie group. The "disappearance of the Huronian" from the geologic terminology might well be postponed until definitive geologic and age studies are available. Penokean Orogeny. The Penokean orogeny resulted in the deformation of a belt of Huronian rocks extending from Minnesota through Wisconsin into :Michigan. The resulting mountains were south of the main belt of the earlier Algoman Mountains, but their general trend was the same. The eastern end of the Penokean orogenic belt, like that of the Algoman belt, is truncated by the Grenville orogenic belt. It is possible, however, that the original Penokean folds, or related belts of folding, extended through the Mistassini district and the Labrador trough (Fig. 1), but the severe metamorphism during the Grenville orogeny at approximately 1.0 b.y. has greatly complicated the geologic record. In central Minnesota, igneous rocks of different ages are distinguished. Tonalites and related rocks were emplaced during an early phase of the Penokean orogeny. The average K-A age for four of these is 1.76 b.y. The St. Cloud Red and the Rockville granites give K-A ages of 1.64 b.y. The indicated time difference between the early tonalites and the younger granites is 120 million years, but results of the Rb-Sr dating, although fewer samples are represented, indicate a time difference of 60 m.y. Neither of the two values may be the absolute time difference, but they set a mini- mum limit of the order of 60 m.y. Three samples of the McGrath gneiss give closely agreeing K-A ages, averaging 1.68 b.y., but a fourth sample gives a lower K-A age of 1.50 b.y. In Wisconsin, a biotite-kyanite schist (KA-45) is dated at 1.72 b.y. and corresponds to the synkinematic McGrath gneiss in the evolution of the Penokean orogeny. Granite near Butternut, Wisconsin, is dated at 1.66 b.y., and a granite dike cutting the folded iron formation in Michigan is -; dated at 1.62 b.y. These are the postorogenic or postkinematic granitic \ intrusions. Because of the cover of glacial drift, the relationships between the rocks DEVELOPMENT OF A PRECAMBRIAN CLASSIFICATION 157 of central Minnesota and those of the Minnesota Valley are unknown. The average K-A age for seven samples of the Montevideo granite gneiss in the Granite Falls-Montevideo region is 1.79 b.y. The K-A ages for the gneiss range from 1.71 to 1.85 b.y. (Table 28). The Rb-Sr ages for two of the samples are lower than the corresponding K-A ages, 1.74 and 1.69 compared to 1.85 and 1.75 b.y. The postkinematic granite (KA-28, 29) in the Granite Falls area is dated at 1.69 (K-A) and 1.65 b.y. (Rb-Sr). The time difference between the synkinematic Montevideo granite gneiss and the postkillematic granite, indicated by the K-A ages, is of the order of 100 m.y., and by the Rb-Sr ages, of the order of 70 m.y. The data for the Montevideo gneiss suggest that the variations in the K-A ages for the gneiss may be greater than the analytical error and may reflect local recrystallization of the biotite, perhaps at the time of intru- sion of the postkinematic granite. It follows also that the ages obtained for the tonalite and related rocks of central Minnesota and for the :McGrath gneiss may be minimum values and not the absolute ages. The Rockville porphyritic granite, for example, is unusual in the development of large crystals of microcline and in the replacement of plagioclase by K-feldspar, and the rock may represent an older tonalite that was altered by potash metasomatism at a later date. Correlation. Radioactivity ages in the range from 1.6 to 1.9 b.y. have been reported from a large number of localities, and some of these date the time of folding of sedimentary sequences accompanied by metamor- phism and granitic intrusion so that a correlation with the Penokean orogeny of the Lake Superior region is indicated. Eckelmann and Kulp (1957, p. 1130) concluded that the pegmatites in the southern part of the Black Hills, South Dakota, were intruded at 1.62 b.y. on the basis of an analysis of a uraninite from the Bob Ingersoll mine. A similar value was previously reported by Wetherill, Tilton, Davis, and Aldrich (1956). Potassium-argon and rubidium-strontium ages (Aldrich, Wetherill, Davis, and Tilton, 1958, p. 1128) for micas from the Bob Inger- soll pegmatite are variable, and the determined ages for lepidolite are lower than the corresponding ages for muscovite. The pegmatites may be considered a late kinematic phase of the orogenic cycle following the depo- sition of a thick sequence of sediments that is well exposed in the northern part of the Black Hills. Collins, Farquhar, and Russell (1954) have summarized the older data from the isotopic lead analyses of samples from uranium deposits in the Churchill geologic province (Fig. 1) in the northern parts of Alberta, Saskatchewan, and Manitoba and the southern part of the Northwest Territories. They date this province at 1.65 to 1.85 b.y. Burwash (1958, 1959) is currently studying the age of the Precambrian basement complex in southern Alberta and southern Saskatchewan and in northern Montana. He finds from K-A dating of biotites that the gneissic complex probably was formed 1.7 to 1.9 billion years ago. 158 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY Some interesting K-A and Rb-Sr ages were recently presented by Was- serburg, Wetherill, and Wright (1959) for minerals from schist and gneiss and from a pegmatite cutting these rocks in Death Valley, California. The determined ages for minerals from the pegmatite are greater than those of the country rock, and this suggests metamorphism to which the schist and gneiss were more susceptible than the pegmatite. The K-A and Rb-Sr ages for the muscovite from the pegmatite range from 1.56 to 1.73 b.y., whereas the ages for the biotite and muscovite from the schist and gneiss range from 0.99 to 1.5 b.y. The ages for the pegmatite suggest an event that possibly may have been contemporaneous with the Penokean igneous ac- tivity in the Lake Superior region. Kouvo (1958, p. 58) has noted the similarity in age of the Svecofennidic and Karelidic orogenic belts of Finland to the post-Huronian or Penokean belt of the Lake Superior region. The Karelidic schist rests on a basement of gneissic granite, and zircon from the Sotkuma granite gneiss of this complex was dated by Kouvo at 2.5 b.y. (Pb 207/206). Biotite from the same rock gave a Rb-Sr age of 1.7 b.y. which is the approximate time of the post-Karelidic orogeny indicated in ages found by Kouvo for zircons separated from intrusive granitic rocks in the Karelidic and the Svecofen- nidic metasediments. The average Pb 207/206 age determined for zircons of two granodiorites and a quartz diorite from the Svecofennidic range is 1.83 b.y. Two oligo- clase granites and a microcline granite from the Karelidic range are dated at 1.80 b.y. The average for six samples of postkinematic granites is 1.65 b.y. The radioactivity ages for the intrusive rocks in the Svecofennidic and Karelidic orogenic belts determined by Kouvo may be compared with the ages from east-central Minnesota and the Minnesota Valley in Figure 36. Tie-lines connect samples dated by different methods. The Rb-Sr and K-A ages for micas found by Kouvo are lower than the corresponding Pb 207/206 ages for the zircons. Kouvo computed the instrumental error of his determinations to be approximately 3 per cent, but he notes that this figure does not include errors arising from incorrect decay constants or from geologic factors. Gerling and Polkanov (1958, p. 717) date the Karelian cycle of sedimen- tation and metamorphism in the eastern part of the Baltic Shield, in Ka- relia and the Kola peninsula, from 1.5 to 1.9 b.y. Synorogenic intrusive rocks are dated from 1.76 to 1.86 b.y., and the postorogenic granites from 1.55 to 1.66 b.y. Greenhalgh and Jeffery (1959) have summarized the ages of uranium minerals and of galen as for Australia. Analyses of da vidite from Crockers Well and from Radium Hill, in the western part of the Willy am a granitic complex in South Australia, give ages of 1.5 to 1.7 b.y. Investigations of the geochronology of the Northern Territory are now in progress at the Massachusetts Institute of Technology (Hurley, personal communication, 1959), and preliminary K-A ages for samples from the Katherine-Darwin 1.5 (/) cr - 1.6 V> 0 W (!) - t:: > i= 1.8 u [,; and 645 to 720 m.y. for Pbeo , IPb"oo. A large number of radioactivity-age determinations have been made by different investigators in a number of laboratories in an attempt to date points in the fossil time scale. The results indicate that revision of the time scale can be expected. The problems include not only the pre- cise stratigraphic assignment of the dated samples but also analytical problems, possible loss or gain of the parent and daughter nuclides, and geologic problems relating to diagenetic and metamorphic history of the samples. Considerable work has been done with glauconite, and a large number of samples ranging from Cambrian to :lVIiocene have been dated by Amirkhanov, Magataev, and Brandt (1957), and Lipson (1958). Three samples of glauconite from Cambrian formations in Minnesota were dated by the K-A method in the present investigation (Table 30). Two samples from the Reno member of the Upper Cambrian Franconia for- mation were prepared by W. E. Crain. A third sample from the .Jordan sandstone was prepared by D. W. Kohls. According to Kohls the glau- conite probably represents reworked material in a silty dolomite near the top of the Jordan sandstone in outcrops along the St. Croix River Sample No. KA-1I3 KA-1I4 KA-192 TABLE 30. K-A AGES FOR GLAUCONITE FROM THE FRANCONIA AND .JORDAN FORNLATIONS K,O K-l1) A'o Description pct. ppm. ppm. A'"/K''' Reno member, Franconia formation .............. 7.87 7.89 0.236 0.0299 Reno member, Franconia formation ..... . ... 7.99 8.01 0.227 0.0284 Jordan formation .......... 7.73 7.75 0.22.5 0.0290 Age m.y. 450 430 43." DEVELOPMENT OF A PRECAMBRIAN CLASSIFICATION 165 in Washington County. Stratigraphically, this sample comes from near the Cambrian-Ordovician contact. The K-A ages (Table 30) agree within the analytical error of the method, and the average age is 440 m.y. Wasserburg, Hayden, and Jen- sen (1956) dated glauconite from the Franconia at Sparta, Wisconsin (Fig. 34), and found a similar K-A age of 435 m.y. Herzog, Pinson, and Cormier (1958) obtained an age of 410 m.y. by the Rb-Sr method for a sample of glauconite from the Franconia formation at Readstown, Wis- consin (Fig. 34). The determined ages for the glauconite samples from Minnesota and Wisconsin agree with the age assignment of the Holmes time scale, but the K-A age of 440 m.y. can be considered only as a minimum age (Goldich, Baadsgaard, Edwards, and Weaver, 1959). There appears to be a difference between the structure and chemical composi- tion of early Paleozoic and Cretaceous or younger glauconites, but factors other than time may be involved. Investigations under way at the Massachusetts Institute of Technology and at other laboratories may resolve some of the problems. A revision of the fossil time scale is suggested by Kulp (1959), who tentatively assigns an age of 540 m.y. to the Lower and Middle Cambrian boundary. Mayne, Lambert, and York (1959), hovvever, favor an age of approximately 650 m.y. for the Upper Cambrian. In addition to the problems of evaluating the radioactivity ages found for the various minerals, the stratigraphic problem of defining the base of the Cambrian is an exceedingly difficult one. The investigations in Minnesota add nothing to the solution of these problems. In Tables 2 and 31 the time boundary between the Precambrian and the Cambrian is shown as 600 (?) m.y., but this is an arbitrary assign- ment to which no special significance should be attached.* As the studies in various laboratories provide new data, it can be expected that atten- tion will be given to this problem. CONCLUDING REMARKS A world-wide classification of Precambrian rocks and an absolute time scale are worthy objectives, and work along these lines should be vigor- ously pursued. The difficulties of describing rocks, of showing their rela- tionships on maps, and of properly interpreting their origin without refer- ence to time have long been apparent to geologists. Experience has shown beyond doubt the futility of distinguishing rock types, such as Lauren- tian, Archean, or Algonkian, on the basis of their appearance, composi- tion, or degree of metamorphism; hence the methods of radioactivity dat- ing are welcomed by geologists. Experience, however, has a second lesson to teach that commonly is * Since this manuscript was completed, Holmes has published "A revised geological time- scale," Trans. Edin. Geol. Soc., 17: 183-216, 1960, in which he places the Precambrian- Cambrian time boundary at 600 million years ago. 166 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY overlooked. In an earlier day the introduction of the petrographic micro- scope revolutionized the study of rocks, and similarly, the mineralo- graphic, the X-ray, and other techniques have had great impact in the various fields of geology. The limitations of these new techniques gradually were determined, and the same must be done for the radioactivity-dating methods. Like the other techniques, the dating methods are powerful and much needed tools, but they cannot stand alone. The so-called classical geological investigations, all too frequently dismissed as being outmoded, are needed as much today as ever before. Classifications, whether they be of minerals, rocks, or stratigraphic units, are in themselves not the ultimate ends of geological investigations, and radioactivity dating has more to contribute to geology than an abso- lute time scale. Read (1955, p. 409), for example, in writing of the origin of granites or of the Granite Series in relationship to mobile belts, states: "Possibly the most creative development in petrogenesis is that based on the realization that a given rock is not formed by chance at an acci- dental time or random place, but that it records precise stages in crustal evolution at definite times at circumscribed places. Petrogenesis has its role in Historical Geology as much as Stratigraphy." Basically the state- ment by Read underlines Historical Geology as an important objective, and exception is not taken to this view, but history which is not well documented in time has a predictable tendency toward becoming garbled. Much work remains to be done in the study of the so-called batholiths of .:\linnesota and Ontario. These are granitic complexes whose origins are related to sedimentary, metamorphic, and magmatic processes spanning a great interval of time. Detailed studies, using a combination of methods, may resolve the time of the Laurentian orogeny which was not accom- plished in the present work. Study of the huge Vermilion complex has scarcely been started, and the same is true of the complex of metamorphic and igneous rocks in east-central Minnesota assigned to the Penokean orogeny. In both of these areas and also in the scattered outcrops of the Minnesota River Valley, it is possible to reconstruct series of events which bear a marked resemblance to the sequences found by many writers in other parts of the world in attempting to relate sedimentary, metamorphic, and magmatic processes to orogeny and mountain building. Particularly useful are the papers by Knopf (1948), Wahl (1949), and Read (1955), whose analyses of the problems reflect their different approaches and their differing emphasis on sedimentary, metamorphic, and igneous processes. The classification of synkinematic, late kinematic, and postkinematic to show the relative position of igneous rocks in the development of an orogenic cycle was suggested by Eskola (1932). Wahl (1936), Simon en (1948), and more recently Marmo (1955a, b) attempted to relate chemical differences in granitic rocks to their positions in the orogenic cycle. The synkinematic intrusions, which Wahl termed synorogenic, are calcic and n;ostIy granodiorites, whereas the late and postkinematic intrusions are DEVELOPMENT OF A PRECAMBRIAN CLASSIFICATION H;7 more typically granites characterized by abundant K-feldspar. The con- cept of the Oberbau or Suprastruktur, as opposed to the Unterbau was introduced by Wegmann (1935). In the present report, suprastructure and infrastructure have been used, although Oberbau has been translated as superstructure. The concept is a useful one. A more detailed program of radioactivity dating in Minnesota and ad- jacent areas might well have as an objective the study of these petro- logical concepts, but detailed mapping and careful chemical analyses, in- cluding the determination not only of the major but also of the trace ele- ments, are badly needed. The combination of methods will afford a more realistic basis for defining an orogenic cycle than is presently available. It is possible that rocks now assigned to an orogenic cycle may be further delimited and assigned to different cycles. Further advances in radioactivity-dating methods through improve- ment of the analytical techniques and refinement of the decay constants may be anticipated. The effects of metamorphism on minerals require ad- ditional study. The K-feldspars from Precambrian rocks show an argon deficiency relative to the associated micas and must be considered to be unreliable for age determination, but A 40 /K40 determinations on feldspar may be useful in indicating metamorphism as was found for the Babbitt and Snowbank Lake areas. Studies are also needed of the retention of argon and of the possible gain or loss of potassium, rubidium, and stron- tium in minerals and in the whole rock. The K-A ages found for the whole- rock samples of slate and phyllite agree, within the analytical error of the technique, with the ages for micas separated from the higher-grade meta- morphic facies of the Thomson formation. The A40/K40 determinations on the whole-rock samples of the granophyres from the Duluth area give ages conforming with geological interpretations, but without additional data from Rb-Sr and U-Pb determinations, the K-A ages of the granophyres cannot be properly evaluated. Compilation of radioactivity ages from different parts of the world shows a marked frequency of dates at 2.6-2.5 b.y., 1.8-1.6 b.y., and 1.1-0.9 b.y. It is unlikely that this pattern is wholly the result of chance. Sub- sidence during sedimentation and the mountain building that follows, as interrelated in the geosynclinal theory, cannot be explained by processes or forces acting entirely within the crust of the earth. Appeal must be made to deep-seated, subcrustal forces, such as convection currents, and with such hypotheses the possibility of major orogenies being essentially con- temporaneous in different parts of the world must be considered. In the present report the geologic names that have been used previously for the Precambrian rocks in the Lake Superior region have been retained, and a plea is made that the names neither be discarded nor formally re- defined until more definitive data are available. Dissatisfaction with our knowledge and present efforts to solve the problems of the Precambrian rocks commonly takes the form of renunciation of the older work. Few 168 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY indeed are the geologic terms, including the time-honored names Archean, Laurentian, Huronian, and Keweenawan, that have escaped severe crit- icism generally, with the recommendation that the name should be rele- gated to the scrap pile of obsolete and worthless nomenclature. The names in themselves may not be of great significance, but there appears to be a danger that in our haste to free ourselves from an imagined bondage of an obsolete terminology we also throw out the geologic method and re- place it with an elusory passe partout. The classification of the Precambrian rocks of the Lake Superior region has gradually evolved through the efforts of many geologists working with the rocks. It is hoped that radioactivity dating in the region will even- tually lead to a better understanding of the temporal relationships of the Precambrian rocks and also of their overall history. The present classifica- tion, summarized in Table 31, is based on the recognition of four periods of marked structural disturbance. From oldest to youngest in Minnesota and adjacent areas, these are (1) Laurentian, (2) Algoman, (3) Penokean, and (4) a phase of the Lake Superior structural disturbance which is essen- tially contemporaneous with the Grenville orogeny of the eastern part of the continent. A possible five-fold division of the Precambrian of the Lake Superior region is indicated in Table 31. For the present, however, the three-fold classification requires no new names and does no great vio- lence to the scheme with which geologists are accustomed to work and may well serve as a basis for further growth. TABLE 31. CO:MPARISON OF lVIAJOR DIVISIONS OF PRESENT CLASSIFICATION WITH OLDER AND POSSIBLE FUTURE DIVISIONS OF THE PRECAMBRIAN OF THE LAKE SUPERIOR REGION Grout et 01, This Report Possible Canadian Future Usage 1951 ERA SYSTEM OROGENY 109 YEARS Revision AGO Paleozoic Paleozoic Paleozoic Cambrian -'0.6 .- ------------ I Grenville ~f-1.0 Late Keweenawan Precambrian II Later Proterozoic Precambrian Penokean ~ 1--1.7 Middle Precambrlon Huronian III Algoman ~f-2.5 Medial Precambrian Timiskomion IV Laurentian - ? - - ----------Early Ontarion Archean Earlier Precambrian V Precambrian ( For de toll s of rock units ond u nconforml ties I see Table 2) APPENDIX, REFERENCES, AND INDEX APPENDIX LOCATION AND DE~CRIPTION OF DATED SAj\lPLES The potassium-argon analytical data for 155 samples and the K-A ages for 134 samples are gi"en in Tables 8, 14, 21. 28, and 30. Twenty-one of the samples are feldspars for which lhe K-A ages are not computed. The analytical data and the Rb-Sr ages for 35 samples are gi"en in Table 7. Of these 31 were analyzed at Minnesota, and .J, in other laboratories. The distribntion or the samples is summarized in Table 32. The locations and descriptions of 126 samples, supplementing information given in the text, are given in the sections of the appendix thal follow. For location of specific cuunties see Figure 37. TABLE 32. NUnlBER A"D DISTRIBUTION OF DATED SAl\lPLES Area K-A Rb-Sr Northern :Minnesota (including Ontario and Mellen, Wisconsin) ............................... 64 19 East-central Minnesota (including Wisconsin and Michigan) ....................................... 30 6 Southwestern Minnesota (including Milbank, South Dakota) ................................... 32 7 Southeastern Minnesota (Cambrian glaue-onites) ........ 3 Miscellaneous ...................................... 5 3 Total . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 134 35 NORTHERN MINNESOTA (PL.l, INCLUDING ONTARIO) The samples in this section are listed in Table 14 and include 15 from On lario and 1 from Mellen, Wisconsin. VERMILION DISTiUCT (A) Soudall-Ely A rea KA-147 Knife Lake slate and graywacke beds at Soudan, Sec. 27 :62-15, St. Louis Co. Loca- tion E. Stop 8, Fig. 14, p. 99 of Guidebook (Klinger, 1956). Dated sample is dark-gray phyllite. X-ray: quartz, muscoyite, chlorite, calcic plagioclase. KA-97 Sericite enclosing a boudin-like mass of red jasper 4 inches long. Soudan mine. crosscut north of main drift on 27th level. Courtesy of F. L. Klinger, Oliver Iron Mining Division, U.s. Steel Corp. X-ray: quartz, muscovite. chlorite. The jasper contains yeins. relaled to fractures. of very fine-grained pyrite ,,·hich also occurs in the sericite. According to Klinger (personal communication, March. 1958), the sample comes from an area that is not close to a known orebody. KA-18 Sericite from the Soudan mine. Massiye sericite from north wall of Shaft Vein ore- body. approximately 25 feet above 27th leyeL Courtesy of F. L. Klinger. Sample comes from within 1 foot of the ore contact. Light-yellowish-gray sericite. X-ray film shows only muscovite. (B) Knife Lake Area KA-200 Knife Lake slate and graY'Yacke. island near west end of Knife Lake, 2 miles easl of the east end of portage from Knife Lake to Seed (I-Ieron) Lake. Dark-gray to black slates interlayered with dark. fine-grained gray,yacke beds which show good graded bed- ding. For chemical analyses of slate and graywacke from this locality see Grout (1933. p. 997). Sample collected by Groul (M2409b). X-ray analysis of dated slate: quartz, mus- covite, chlorite, calcic plagioclase. C02, 0.03 per cent (SSG); C, 0.12 per cent (C. O. Inga- mells). 171 172 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY FIGURE 87. - Counties o f Minnesota. KA-228 Knife Lake slate, south shore of Knife Lake, approximately 0.6 miles west of portage to Eddy Lake, SW'\4 Sec. 20:65-6. From an outcrop mapped as an iron-bearing lens in the Knife Lake (Gruner, 1941 , PI. 1). Dark-gray, slickensided argillite . Thin sec- tion shows mostly feldspar , with quartz, hornblende, minor sericite; shards indicate pyro- clastic origin, a crystal-vitric luff. KA-244 Knife Lake slate, just north of Eddy Lake and above waterfalls, along portage from Eddy to South Arm of Knife Lake, SEl", Sec. 20:65-6. Da rk-gray sla te or argillite with poor secondary cleavage. X-ray : quartz, muscovite, chlorite, calcic plagioclase. KA-243 Knife Lake graywacke-slate, a long portage from J ean (Zeta) Lake to Eddy Lake. NEJ4 W J4 Sec. 28: 65-6. Dark-gray, fine-grained graywacke with subconchoid al fracture. X-ray: quartz, muscovite, chlorite, calcic plagioclase. KA-242 Knife Lake slate, along portage from Ogishkemuncie Lake to Annie (Dike) Lake. APPENDIX 173 NW1,4SWlj! Sec. 27:65-6. Dark-gray slate. X-ray: quartz, muscovite, chlorite, plagio- clase. KA-240 Knife Lake slate, south side of southernmost point on the east side of Cypress Lake, near middle of south line Sec. 23: 66-6 (Fig. 13). Fine-grained, brittle rock, breaks with conchoidal fracture, producing sharp, glass-like edges. X-ray: quartz, chlorite, mus- covite, plagioclase. (C) Snowbank Lake A rea (Fig.H) KA-I06 Syenite, coarse-grained, with gray and pink feldspar, hornblende and biotite. From the southeast shore of Snowbank Lake, SE% Sec. 31:64-8, Lake Co. Collected by C. W. Sanders (1929, M2450). KA-323 Knife Lake conglomerate. SE cor. Sec. 29: 64-8, opposite north end of long island in Snowbank Lake. Collected by C. W. Sanders (1929, M2449). Biotite from selvage around 3-inch cobble of dark-gray diorite. KA-229 Microgranite porphyry with phenocrysts of light-pink feldspar. North side of Snowbank Lake. NW1.4 Sec. 2(j: 64-9, Lake Co. ( D) Saganaga Lake Area ( Fig. 13) KA-328 Pegmatite on Gold Island in southwestern part of Saganaga Lake at the head 01 Red Rock Bay. Red K-feldspar and light greenish-yellow sericite collected from dump. representing material taken from a prospect hole on the northwest side of island. CANADA (EJ Ontario (Pl. :; j KA-76 Northern Light gneiss from north side near east end of Moose Island, Northern Light Lake (see Fig. 13). Light-gray, fine- to medium-grained biotite gneiss interlayered with dark-colored schist. Paragneiss: quartz, 30; plagioclase (Anu), 55; microcline, 10; biotite, 5 per cent. KA-77 Northern Light gneiss from small island near east shore, southern part of Sayage Bay, Northern Light Lake (Fig. 13). Light-gray, fine-grained, banded paragneiss similar toKA-76. KA-262 Biotite granite gneiss, Highway 17, 7 miles east of English River and iust south of Sheba station on Canadian Pacific Railway. Well-developed foliation strikes N. 85° E.; dips 65° S. Light-gray, medium-grained paragneiss with coarse-grained schlieren of bio- tite; cut by numerous dikes of granite, pegmatite, and aplite. Modal analysis by J. K. Frye: quartz, 25; oligoclase (An,.), 70; K-feldspar, 3; biotite, 2 per cent. Collected by D. H. Yardley. KA-112 Quartz-biotite schist, Briarcliffe Lake iron prospect, diamond drill hole. F 115 at 260-foot depth, just east of Melchett Lake, 37 miles northwest of Nakina, Ontario. Fine- grained schist or graywacke gneiss containing minor intermediate plagioclase and garnet. Iron formation and schist strike N. 70° E.; essentially vertical. Collected by E. R. Mead, courtesy of Anaconda Company (Canada), Ltd. KA-l11 Quartz-plagioclase-muscovite pegmatite, Briarcliffe Lake iron prospect, diamond drill hole. F 113 at 340-foot depth, Nakina, Ontario. Coarse-grained, light-colored pegroa- tite cutting schist (KA-IU). Collected by E. R. Mead, courtesy of Anaconda Company (Canada), Ltd. KA-258 Granite from west side of small island near east shore of Kashabowie Lake. Ontario. Medium-grained, leucocratic; contains inclusions of schist. Modal analysis by J. K. Frye: quartz, 26; oligoclase (An16), 38; microcline, 31; biotite and others, 5 per cent. (226") ada- mellite. Collected by D. H. Yardley. KA-298 Granite, composite of three samples of Algoman granite north of Eye Lake, west of Steeprock Lake. Coarse-grained, pink. Estimated composition: quartz, 20; oligoclase, 45; microcline, 30; biotite and others, 5 per cent. Courtesy of W. J. Huston, Steep Rock Iron Mines Ltd. KA-21 Lamprophyre dike, Cold stream Copper Mines, approximately 8 miles southwest of Kashabowie and % mile northeast of Burchell Lake, Thunder Bay District, Ontario. Dikes cut Keewatin complex; composition: biotite, 52; andesine (An3'), 26; calcite, 15; quartz, minor chlorite, and pyrite, 7 per ('ent. determined by H. L. Taylor (unpublished M.S. thesis, Univ. Minn., 1957). Collected by Taylor. KA-191 Coutchiching paragneiss, from north shore of entrance to upper Rice Bay, Rainy 174 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY Lake, Ontario. Outcrop is cut by lamprophyric and aplitic dikes, Mode by Z. K Peterman: quartz, 38; albite, 39; biotite, 21; muscovite, 1; microcline, 1 per cent. KA-178 Porphyroblastic biotite-epidote gneiss, east side of largest island in Rocky Islet Bay, Rainy Lake, Ontario. Mapped by Lawson (1913a) as Algoman syenite gneiss. Dark- gray, mIcaceous paragneiss with abundant large porphyroblastic microcline. Estimated composition: quartz, 20; calcic oligoclase, 30; microcline-perthite, 25; biotite, 15; epidote, 10 per cent. Numerous granitic or aplitic dikes. KA-320 Aplite dike cutting Coutchiching paragneiss, south side of upper Rice Bay, mapped by Lawson as Laurentian granite. Gray, fine-grained; modal analysis by Z. K Peterman: quartz, 35; oligoclase (An,,), 42; microcline, 12; muscovite, 6; biotite, 3; epidote, 1 per cent. Leucogranodiorite. KA-'222 Granite, Ben Island, soutb of Rice Bay, Rainy Lake, Ontario, mapped by Lawson as an Algoman intrusive in the Coutchiching. Gray to pink, medium-grained, with local well-de\'eloped foliation. Estimated composition: quartz, 22; microdine, 43; albite-oligo- clase, '25; biotite, 5; muscovite, 5 per cent. KA-224 Granite gneiss, small island south of Goose Island, mapped by Lawson as AJgoman syenite g1.eiss. This rock is considered a border phase of the Algoman granite which in- trudes the Coutchiching. Dark-gray, medium-grained, with distinct foliation; contains quartz, feldspar, and '20-'25 per cent of biotite. KA-'272 Granite gneiss, Highway 17, 1.5 miles north of Emerson Lake, east of Lake of 'Voods, 5 miles east of junction of highways to Kenora and to Fort Frances. Medium- grained, dark-red, biotitic gneiss; foliation strikes N. 70° K, attitude essentially vertical. Estimated composition: quartz, 20; K-feldspar, 15; olig'oclase near albite, 40; biotite, 15; epidote, with minor sericite, chlorite, leucoxene, etc., 10 per cent. Collected by D. H. Yardley. I\:A-70 Lamprophyre dike cutting Saganaga granite, black, coarse-gTainerl, biolitic. Sam- pled at northern end of dike on small island in Ontario, approximately 1 mile north of J'.Iinnesota-Ontario boundary (Fig. 13). VERMILION GRANITE REGION (P) 1'ermilion Granitic Complex KA-47 Vermilion granite. Biotite schlieren from granite of border phase of the Vermilion complex, from west end of Levorsen's Island, Burntside Lake, NE% Sec. 18: 63-12. Col- lected bv D. All. KA-,t6 ,; ermilion granite gneiss, blasted outcrop, east side of Echo TraiL just sonth of junction with road to Everett Lake, NE14 Sec. 31: 64-12, St. Louis Co. Granite gneiss amI migmatile. The regional structure trends E-'V and was produced by granitization of a sedimentary sequence or by injection of granitic magma, probably a combination of the processes. Plastic deformation has resulted in crenulated folds which resemble ptygmatic folds, except that the axial planes are parallel to east-west foliation. Boudin- age is well developed. The tightly folded material and the boudins are composed of quartz and white feldspar with little or no biotite. Biotite commonly is concentrated all the margins. Differential melting may have produced the leucocratic granite in the highl)' deformed structures. A north-south dike cuts the gneiss with the east-west folia- tion. The dike is deformed, suggestive of plastic or rheomorphic deformation. KA-239 Vermilion granite. Black, biotite-rich schlieren in gray granite blasted in high- way construction. Echo Trail. 8 miles east of intersection with Crane Lake-Orr road. Se~ .. 2: 65-16, St. Louis Co. KA-249 Vermilion granite gneiss in highway cut, 6.6 miles north of Cusson, Sec. 22:66-20, St. Louis Co. High-grade metamorphism (garnet-amphibolite facies) of a sedimentary sequence, originally composed of graywacke and shale. Light-gray, coarse-grained granite gneiss composed of quartz, microcline, oligoclase, with biotite-rich layers and schlieren. Crenulated folds and boudinage conspicuously developed (see Fig. 15). Some layers of the g11eiss contain abundant garnet. Gneiss is intruded by a younger pink-colored granite. KA-250 Graywacke gneiss. Blasted outcrop north side of U.S. Highway ,53, at intersection ,,·ith State Highway 73, middle ,,,est line of Sec. 19:63-19. Principal rock type is dark- gray, fine- to medium-grained graywacke gneiss, striking N. 85° W. and dipping 70° S. Quartz boudinage developed from original quartz veins. Some granite gneiss at west end of outcrop. KA-fl4 Pegmatite, Net Lake. Outcrop in w(>ods, north of road. approximately 1/1 mile east of Net Lake, on outskirts of village. SE14 Sec. 18:65-21, St. Louis Co. Black biotite schist, APPENDIX 170 some very coarse-grained, cut by white sinuous pegmatite dikes, 4-6 inches across. Biotite in variously oriented books with gray interstitial material composed of sericitized plagio- clase, epidote, and apatite. Collected by D. Alt. .. . . . KA-4 Pegmatite. approximately 4 miles northwest of Kmlllouul statlOn and iust east of Duluth, Winnipeg. and Pacific RaihnlY. Prospect pit, 50 f~et east of center lme and 2.50 feet south of north line of SW1,4SElJ. Sec. 9:67-21, St. LoUIS Co. Coarse, black and white biotite-oligoclase (An,,) pegmatite. Collected by G. M. Schwartz. KA-83 Graywacke gneiss, small outcrops exposed by highway construction along U.S. Highway 53, approximately 1.5 miles south of International Falls, Koochiching Co. Light- gray, fine-grained gneiss, with lenses of coarse-grained quartz. KA-S2 Granitic gneiss, outcrop on north side of State Highway II, % mile northwest of Birchdale, Sec. 33: 160-27, Koochiching Co. KA-307 Tonalite, from the south prong of the bay at the northeast end of Grassy Island. mapped by Lawson (1913a) and Grout (1925a) as Laurentian; as Algoman by Cram (1932) and by Grout et al. (1951). Grayish-pink. fine- to medium-grained, massive rock. Modal analysis by J. K. Frye: quartz, 23; plagioclase (An,,;), 58; microcline, 4; biotite, 10; a('c&~­ sory and alteration minerals, 5 per cent. (G) N01·thwest Angle (Pl. 3) KA-17 Rader pegmatite, Sec. 6: 167-33, Lake of the Woods Co. (Northwest Angle), Minne- sota. Pegmatite was once mined by W. C. Rader for feldspar which occurs in large crystals, I foot or more in length. Feldspar is white albite, but there is also pink microcline-perthite. Both feldspars are intergrown with quartz. :Muscovite is in large books, but most is not of commercial grade. Beryl is in well-formed crystals, as much as 8 inches across the hexag- onal section. Some garnet occurs in small crystals. ~Iuscovite sample was collected by Grout. KA-88 Pegmatite, exposed in small uranium-prospect pit near the center of Sec. 32: 168-34, approximately 4.5 miles west of the Rader pegmatite. Large crystals of microcline-perthite, oligoclase, quartz. and biotite. Country rock is granitic gneiss and schist; high-grade meta- morphic rocks show relic folds. GIANTS RANGE REGION (H) Giants Range Granitic Complex KA-236 Grayish-pink porphyritic granite, in field east of County Road 38, two miles north of Grand Rapids, NW cor. Sec. 4: 55-25, Itasca Co. Approximate composition: plagioclase (An,.), 35; quartz. 30; microcline, 30; biotite and accessories. 5 per cent. Porphyroblasts of microcline include myrmekitic plagioclase, quartz, and biotite. Alteration products: chlorite, sericite, epidote. KA-l!37 Dark-gray, coarse-grained gneiss with compositional banding and layers rich in biotite. Small outcrops in woods, approximately I mile northeast of Nashwauk, NW% Sec. 21 :57-22, Itasca Co. Granoblastic. sutured grains, principally oligoclase with minor micro- cline. hornblende, epidote; granodiorite or tonalite. KA-48 Granite from quarry north of Mountain Iron, NW14SE% Sec. 2S:59-18, St. Louis Co. Medium-grained pink gneissic granite, containing xenoliths and schlieren. KA-105 Granite gneiss from E1f:JNE1,4 Sec. 28 :59-17, St. Louis Co .. approximately 3 miles northeast of Virginia. Collected by I. S. Allison (M2865). KA-317 Gneiss. ontcrops on west side of U.S. Highway 53, approximately 5 miles north of the city limits of Virginia, Sec. 18:59-17. St. Louis Co. Dark-gray coarse. foliated or rudely banded gneiss, with foliation striking E-W. Cnt by yonnger pink granite; 10 miles north of this locality, roadcuts show fine- to medium-grained gray granite which is intruded by pink granite and tourmaline-bearing, pegmatitic granite. KA-316 Gneiss, roadcut along U.S. Highway 53, approximately I mile north of Idington, in the SW14 Sec. 21 :61-1S, St. Louis Co. Approximately at contact of Giants Range gran- ite and older schist and gneiss (Allison. 1925). Medium-grained gneiss composed of plagio- clase, quartz, and about 30 per cent of hornblende and biotite. Outcrops appear to be largely graywacke gneiss; leucocratic dikes deformed to boudinage. KA-81 Granite, second large outcrop along State Highway 1. 2.7 miles south of the inter- section of Highways I and 169 in Ely, along the west line of Sec. II: 62-12, St. Louis Co. Contaminated phase of the Giants Range granite: black and pink. medium-grained horn- blende-biotite tonalite or granodiorite. This loeality is I mile west of White Iron Lake where there are numerous inclusions of greenstone and other rock types in the granite. 176 PRECAMBRIAN GEOLOGY AND GEOCHRONOLOGY KA-8 Granite. outcrops of the Giants Range granite overlain by the Biwabik formation. Reserve Mining Company property. former town of Babbitt, NW14, Sec. 17:60-1~ and in adjacent sections. Coarse-grained, pink hornblende granite with some biotite. KA-137 Virginia argillite. cored sample from drill hole No. 8519, ~O-Q.5 feet, Virginia, Sec. 5 :58-17, St. Louis Co. Fine-grained, black, graphitic argillite. R. W. Marsden, courtesy or Oliver Iron Mining Division, U.S. Steel Corp. KA-Ql~ Virginia argillite, cored sample from drill hole No. 3961, 16~-171 feet, southwest of Grand Rapids, SW14NE14 Sec. 4:54-~6, Itasca Co. Courtesy of R. W. Marsden. Fine- grained black argillite. CO" 1.1~ per cent (SSG); C, 7,90 per cent (C. O. Ingamells). For a complete chemical analysis of the graphitic argillite of the Virginia formation from this area see White (1954, p. 17). NORTH SHORE REGION (1) Duluth Complex KA-295 Pyroxenite, from near the base of the gabbro in Duluth, upper tracks of tilt' Canadian National Railroad, SW14 Sec. 33: 49-15, St. Louis Co. Greenish-black, eoarse- grained biotite pyroxenite. KA-9~ Diabase, from the Beaver Bay complex. mouth of Beaver River, Sec. 12:55-8. Lake Co. Collected and biotite sample prepared by H. M. Gehman, Jr. KA-176 Ophitic diorite, diamond drill core from hole drilled in the vicinity of Aurora. Coarse-grained biotite in dark coarse-grained rock composed of andesine, pyroxene, bio- tite, ore minerals. Courtesy of Donald ·Wager. KA-65 Gabbro, Duluth gabbro from D. D. H. No. 235, NE cor. Sec. ~0:60-1~, Babbitt. Biotite sample prepared by J. R. N. Gundersen, courtesy of Reserve Mining Company. KA-296 Pegmatite, in Biwabik iron-formation at Babbitt, from main pit in NW14SW14 Sec. 16: 60-1~. Large flakes of green biotite associated with albite and quartz. Sample col- lected and prepared by J. R. N. Gundersen. (J) Contact Rocks Related to Duluth Gabbro KA-23~ Thomson formation, north of Grandview golf course, west of Duluth, east side of Midway Road and approximately 1_14 miles north of U.S. Highway 61, in NWl,4SW14 Sec. 17:49-15. Outcrops contain graywacke and slate; strike N. 85° E.; dip, 80-85° S.; approximately '%, mile west of the base of the gabbro. Sample KA-Q32 is fine-grained graywacke-slate. X-ray: quartz, biotite, amphibole, plagioclase. Collected by F. F. Grout. KA-233 Thomson formation from same locality as KA-232. Dark-gray, well-indurated slate or graywacke-slate. X-ray: quartz, biotite, plagioclase. Collected by F. F. Grout. KA-131 Rove formation, roadcut along the Gunflint Trail, south of the west end of Gun- flin t Lake, in NW%SW% Sec. 30: 65-3, Cook Co. Dark-gray, hard, recrystallized argillite with a subconchoidal fracture, at contact with diabase sill. X-ray: quartz, biotite, amphi- bole, plagioclase. KA-177 Contact rock, diamond drill hole in vicinity of Aurora. Light-gray, fine-grained sugary, with abundant fresh, black biotite flakes. Apparently a granoblastic contact rock probably derived from a sediment. Courtesy of Donald Wager. KA-8 Granite at Babbitt. see Section H, above. KA-I06 Syenite at Snowbank Lake, see Section C, above. KA-93 Granite, Mellen, Wisconsin, medium-grained, red granite from SW cor. Sec. 24, T. 45 N., R. 3 W., Ashland Co. Collected by George A. Moedein, and described as post- gabbro granite. Courtesy of P. A. Bailly, Bear Creek Mining Company. (K) Duluth Area Granophyres KA-I0l Granite dike in dellenite flow, 8th Street at 3rd Ave., Duluth, Sec. ~7:50-14. (Goldich, Taylor, and Lucia, 1956, modal analysis No. ~8, Table 5, p. 81.) Chemical analysis No.4, Table 17. KA-I02 Granophyre intrusive near top of anorthositic gabbro, quarry on Woodland Ave" Duluth, SE14SWl,4 Sec. 2: 50-14. (Taylor, 1956, modal analysis No.1, Table 6, p. 60.) KA-I03 Granophyre intrusive in basalt flows above the gabbro complex, quarry on Ken- wood Ave., Duluth, Sec. 15: 50-14. (Taylor, 1956, modal analysis No. ~, Table 6, p. 60.) KA-104 Gl'anophyre intrusive along the contact of the layered series and the anorthositic gabbro. 27th Ave. at 12th St., Duluth. APPENDIX 177 EAST-CENTHAL :MINNESOTA (PL. 't, INCLUDING WISCONSIN AND lVIrCHIGAN) The 31 samples in this section are included in Table 21. Twenty-two are from east-central Minnesota; 9 from Wisconsin and Michigan. (A) Thom_~on Formation Metasediments KA-3.5 Dark-gTay to black slate with well-developed secondary cleavage. from qnarry, Sec . .5: 48-16, just south of reservoir and west of Thomson, Carlton Co. KA-38 Gray phyllite. Sec. 24: 48-18, 0.4 mile west of Atkinson, on north side of road. Carlton Co. KA-39 Dark-gray. crinkled phyllite. NW cor. Sec. 1: 46-19. Cut along Northern Pacific Railway at station in Barnum, Carlton Co. KA-40 Medium- to coarse-grained schist. Locally the micaceous minerals are wrapped around pod-like masses of milky quartz. NE1,4 Sec. 20: 46-19. Cuts along Minneapolis S1. Paul Sault Ste. Marie Railroad, north of station in Moose Lake, Carlton Co. KA-96 Gray, quartz-mica phyllite from outcrop on north side of Little Elk River, approxi- mately 2 miles north of Little Falls, SW cor. SE1,4 Sec. 31: 130-29, Morrison Co. (B) McGrath Gneiss KA-41 Coarse-grained gneiss, bandecl with black micaceous and light-gray quartz-feldspar layers. Prominent porphyroblasts of red K-feldspar. SE14 Sec. 21: 4.5-21. Cut along Min- neapolis St. Paul Sault Ste. Marie Railroad, 2 . .5 miles west of Denham, Carlton Co. KA-43 Pink. coarse-grained gneiss. distinctly foliated but not banded. Large pink K-feld- spar, locally pegmatitic. SW cor. Sec. 34: 45-21.3 miles southwest of Denham, Carlton Co. KA-164 Coarse-grained gneiss. banded with biotite-rich and quartz-feldspar layers. SW1,4 SW14 Sec. 1'43-24, approximately 2 miles southwest of McGrath. Aitkin Co. KA-63 Pink and gray, coarse-grained gneiss with distinct foliation and some bancling. Quarry prospect. NW1,4 Sec. 6: 44-23. 2 miles west of Dads Corner. Aitkin Co. (C) !II etasediments from the Cuyuna Di,