ENGINEERING RESEARCH INSTITUTE UNIVERSITY. OF MICHIGAN ANN ARBOR STUDIES IN THE NATURAL HISTORY OF MICAS SEPTEMBER 1951 - AUGUST 1953 By E. Wm. Heinrich A. A. Levinson D. W. Levandowski C. H, Hewitt Project M978 CHEMICAL-PHYSICS BRANCH, SQUIER SIGNAL LABORATORY UNITED STATES SIGNAL CORPS, FORT MONMOUTI, N. J. DA 36-039 sc-15357, SC PROJECT NO. 152B-0, DA PROJECT NO. 3-99-15-022 Under the administration of' S. Benedict Levin, Deputy Chief August, 1953

TABLE OF CONTENTS Page LIST OF TABLES viii LIST OF FIGURES xii ABSTRACT xiii PART I. GENERAL INFORMATION I. INTRODUCTION 1 A. General Statement 1 B. History and Personnel of the Project 3 C. Acknowledgements 4 II-. CRYSTAL CHEMISTRY OF MICAS 7 III. X-RAY TECHNIQUES FOR IDENTIFICATION OF POLYMORPHS 10 IV NOMENCLATURE 11 Vc CLASSIFICATION 26 A. History of Mica Studies 26 B. Chemical-Structural Systematization of the Natural Micas 34 PART II. MUSCOVITE I. MINERALOGY OF NORMAL MUSCOVITE 44 A. Chemistry 44 B. Muscovite Structure 45 1. General and Polymorphism 45 2. Morphology 50 3. Interferometric Data 55 C. Optical Properties 57 1. Indices, 2V 57 a. Variation of optical properties with composition 57 b. Changes of optical properties with temperature 58 2. Variation of Color with Other Optical Properties and Composition 59 D. Zoning 61 II. MINERALOGY OF VARIETAL MUSCOVITES 66 A. Oellacherite and Other Reported Barium-Containing Micas 66 B. Ferrian and Manganian Muscovite 69 C,. Chromian Muscovite 71 iii

TABLE OF CONTENTS (cont.) Page D. Phengite 74 lo Introduction 74 2. Alurgite 76 3. Mariposite 78 4. Phengite 79 5. Muscovite, Sultan Basin, Washington 80 6. Other Reported Uniaxial or Near-Uniaxial Kuscovites 80 7o Discussion 82 E. Lithian Muscovite 85 III o Occurrence 91 A. Granites and Related Rocks 91 B. Pegmatites 91 1. General 91 2. Deposits 92 a. Disseminated books in unzoned pegmatites 92 b. Wall-zone deposits 92 c. Intermediate-zone deposits 93 d. Central-unit deposits 93 e. Fracture-controlled veins and replacement deposits 94 fo Combined deposits 94 35 Exomorphic Mica 95 Co Hydrothermal Deposits 95 D. Metamorphic Rocks 95 E. Sedimentary Rocks 96 Fo Variation of Composition with Occurrence 96 IV. Pegmatitic Muscovite 98 A. Structural Defects 98 o1 Muscovite-Biotite overgrowths and Intergrowths 99 B. Inclusions 104 1. Magnetite and Hematite 104 2. Other Minerals 104 3. Other Materials 104 C. Variations in Individual Pegmatites 104 1o Concentrations 104 2. Structural Defects and Color 105 3. Composition 106 D. Variations in Pegmatite Districts 118 1. Occurrence and Color 118 2. Composition 118 E. Commercial Quality Factors 129 iv

TABLE OF CONTENTS (cont.) Page PART III. LEPIDOLITE I. MINERALOGY 153 A. Composition 130 B. Structure 131 1. Polymorphism 131 2. Single-Crystal Variations 133 3. Uniaxial Lepidolite 135 4. Powder X-Ray Studies 138 5. Variation of Structure with Composition 139 6. Discussion 142 7. Summary 146 C. A New 3-Layer Monoclinic Lepidolite Polymorph 147 D. Optical Properties 150 1. Indices, 2V, and Variation with Composition 150 2. Variation of Color with Composition — 154 II. OCCURRENCE 160 A. Non-Pegmatite 160 B. Pegmatites 160 1. Localization 160 2. Macrostructural Features 161 a. General 161 b. Overgrowths 161 3. Compositional Variation in Individual Pegmatites and Districts 164 PART IV. PHLOGOPITE I. MINERALOGY 169 A. Composition 169 1. General 169 2. Mahadevite 169 B. Structure 170 1. Polymorphism 170 2. Relation of Structure to Composition 172 C. Optical Properties 172 D. Inclusions 178 E. Zoning 178 II. OCCURRENCE 182 A. Non-Pegmatitic 182 B. Pegmatitic and Vein Occurrences 182 III. MANGANOPHYLLITE 183 A. Introduction 183 B. Structure 183 C. Composition 183 D. Optics 184

TABLE OF CONTENTS (cont.) Page Eo Single Crystal Variation 184 F. Occurrence 185 Go Conclusions 185 PART V. BIOTITE I. MINERALOGY 187 Ao Composition 187 B. Structure 187 1. Introduction 187 2. X-Ray Technique 188 3, New Structural Data 189 4. Variation of Structure in Pegmatite Districts 189 C. Optical Properties 191 1o Indices, 2V, Optic Plane; Variation with Composition 191 2. Color; Variation with Composition 191 3. Zoning 192 II. OCCURRENCE 194 Ao Igneous Rocks 194 B. Metamorphic Rocks 194 C. Variation of Composition with Occurrence 194 III PEGMATITIC BIOTITE 198 A. Localization 198 B. Macrostructural Features 199 PART VI. ZINIWALDITE I. MINERALOGY 200 A, Composition 200 B.o tructure 201 Co Optical Properties 204 Do Zoning and Overgrowths 205 II. OCCURRENCE 208 PART VII. MISCELLANEOUS MICAS I. PARAGQNITE 211 II. ROSCOELITE 212 IIIo TAENIOLITE 216 vi

TABLE OF CONTENTS (cont.) Page PART VIII. PARAGENESIS OF PEGMATITIC MICAS I, SEQUENCE OF CRYSTALLIZATION OF MICAS 217 Ao Zonal Growths 217 B. Overgrowths between Micas 219 C. Lpcalization in Pegmatites 219 II. CHEMICAL EVOLUTION 225 REFERENCES 229 vii

LIST OF TABLES Table Page I. New Spectrochemical Data on 168 Muscovites.....o................. 45 II. Peacock and Ferguson (1943) Measurements on Muscovite o......... 53 III. Variation of Color with Composition in Muscovite (New Data)...... 60 IV. Optical Data on Selected Zoned Muscovites.................o....o 63 Vo Spectrochemical Analysis of Mangan-Muscovite (Eskola, 1914)...... 71 VI. Phengites with the Two-Layer Structure, Analyzed by Jakob (1925B, 1929B, 1929C).............o........o............... * 79 VII. Occurrences of Uniaxial or Nearly Uniaxial Phengites and Muscovites o.o...eooooo.......................................oooe...o... 84 VIII. Approximate Observed Intensities of Some (Ok,) Reflections of Normal Muscovite and Lithian Muscovite,....o.o.............. 88 IX. Spectrochemical Analyses of Muscovite Sheets from Opposite Ends of a Single Thick Book. (Tilley Pegmatite, FranklinSylva District, N. C.).........................o........... 106 Xo Spectrochemical Analyses of Different Muscovites from the Same Zone Within the Big Ridge Pegmatite (Franklin-Sylva District, No C.)OO.oe eo....ooo.......oo...........oooo............ 107 XI. Spectrochemical Analyses of Different Muscovites from the Same Zone within the Poll Miller Pegmatite (Franklin-Sylva District, N. C.)................... o. o o oo a..a..... 108 XIIo Spectrochemical Analyses of Different Muscovites from the Same Zone within the Tilley Pegmatite (Franklin-Sylva District, No C. )O o o * 9 0 0 0 O a O.. a o. o..... o o o a...o a.. o 0 0 0 a *.o * o.*. 109 XIII. Spectrochemical Analyses of Different Muscovites from the Same Zone within the Ruby, tNorton, Sheep Mountain and Doc Nicols Pegmatites, 7( Franklin-Sylva District, N. C.)*.......... 110 XIV. Spectrochemical Analyses of Different Muscovites from the Same Zone within the Mitchell Creek and Johnson Pegmatites (Thomaston-Barnesville District, Georgia).....a................ 111 Vili

LIST OF TABLES (cont.) Table Page XV. Spectrochemical Analyses of Different Muscovites from the Same Zone within the School Section Pegmatite (Eight Mile Park, Colorado)......................................... 112 XVI. Spectrochemical Analyses of Different Muscovites from the Same Zone within the Meyers Quarry Pegmatite (Eight Mile Park, Colorado)..........................o............. 113 XVII. Spectrochemical Analyses of Different Muscovites from the Same Zone within the Globe Pegmatite (Petaca District, New Mexico).................................................. l114 XVIII. Spectrochemical Analyses of Muscovites from the Gregory Pegmatite (Franklin-Sylva District, N. C.)....................... 116 XIX. Spectrochemical Analyses of Muscovites from the KXier Pegmatite (Franklin-Sylva District, N. C.).0......................... 117 XX. Minor and Trace Elements in Muscovites from Pegmatites (Franklin-Sylva District, North Carolina and North Georgia)...........................................o.............. 120 XXI. Minor and Trace Elements in Muscovites from Pegmatites (Spruce Pine District, North Carolina)....................... 121 XXII. Minor and Trace Elements in Muscovites from Pegmatites (Shelby-Hickory District, North Carolina)...,,.................... 122 XXIII. Minor and Trace Elements in Muscovites from Pegmatites (Thomaston-Barnesville District, Georgia)..0..................... 123 XXTV. Minor and Trace Elements in Muscovites from Pegmatites (Alabama District )........................... 124 XXV. Minor and Trace Elements in Muscovites from Pegmatites (Latah County District, Idaho).................0.................. 125 XXVI. Minor and Trace Elements in Muscovites from Pegmatites (Petaca District, New Mexico)...................................... 126 XXVII. Minor and Trace Elements in Muscovites from Pegmatites (Eight Mile Park District, Colorado)............................. 127 XXVIII. Minor and Trace Elements in Muscovites from Pegmatites (Southwestern Montana),**.................. **...*.... 128 ix

LIST OF TABLES (cont.) Table Page XXIXo New Spectrochemical Data on 26 Lepidolites o...................... 132 XXXo Structure of Micas Analyzed by Stevens (1938)..... o...e....... 134 XXXI. Extent of Polymorphic Variation in One Lepidolite Book from Opportunity Pegmatite, Gunnison County, Colorado.....o.......o... 136 XXXIIo Spacings of Polymorphic Forms,.o.,o.....o ooo....oo........ 141 XXXIII. Structure of Micas Described by Berggren (1940, 1941) and Lundblad (1942 ) o..................e......oo.. oeo o o o.. o e o o 14 XXXIV. Spectrochemical and X-Ray Data on Lepidolite from Skuleboda, Sweden (No. 476)...oo o.....,...o o o o o.. o............... 151 XXXV. Indices of Refraction and 2V of Lepidolites and Lithian Muscovite O o..o o o.o o... o o oo o o...o o o..o oo * 153.... 155 XXXVI. Relationship between Color and Composition in Lepidolites, New Analyses.........o. o.e e *. * o *,...o. 156 XXXVIIo Relationship between Color and Composition in Lepidolites, Reanalyses of Lepidolites Reported in the Literature.1........ 158 XXXVIIIo Relationship between Color and Composition in Lepidolites; Specimens in Our Laboratory for which Analyses are Recorded in the Literature o oo.,.ooo.o...,ooo,,oooooooooooo ooooooooooo Oo 159 XXXIXo Analyses of Lepidolites from Brown Derby Pegmatite, Gunnison Coulnty, COlOrad1o~..o o..<. e.... oebeoooeoooo oo oeooooe o.o.~1 65 XL. Spectrochemical Analyses of Micas from Newry, Maine......,....... 166 XLIo Reanalyses of Two Lepidolites from Varutrask, Sweden Analyzed by Berggren o.o e o...o.eo.o o O O O O O O O a O O O O. O O o....... O........ 167 XLII. Spectrochemical Analyses of Lepidolites from Two Other Pegmatites in Brown Derby Districto8ooo..., oeoooo.. e.............. 168 XLIIIo 2- and 3-Layer Phlogopiteso....o.o.......o............ 171 XLIV. Structure of Analyzed Biotites, Phlogopites and ManganophylliteSoooooo. o. o o o o a a a o a a 173 XLV. Spectrochemical Analysis of Light-Green Phlogopite from Snake Creek, Utaho e,.. o e e o.. d o.o. o o o o o o oo o o Q o o o o o o o o o o o o. * O 179 XLVIo Spectrochemical Analysis of ZonedPhlogopite, Perth, Ontario. 179 XLVII. Occurrence of Manganophyllitee.........,.......... 186 XLVIII Structure of Biotite Varieties oo o o o o o o o,.. o o,.. o o o........ 190 x

LIST OF TABLES (concl.) Table Page XLIX. Chemical Relati6n of Zinnwaldite to Lepidolite and Biotite..... 200 L* New Structural Studies of Zinnwaldite............................ 202 LI.- Approximate Observed Intensities of Some (Ok;) Reflections of 2-Layer Forms....................... 204 LII~ Optical Properties of Zinnwaldite (from the Literature).......... 206 LIII. Occurences of Zinnwaldite....0.............................0 *.. 209 LIV. X-Ray Powder Data for Roscoelite and Barium Muscovite........... 214 LV.6 Spectrochemical Analyses of Muscovites from Jasper Pegmatite, (Franklin-Sylva District, N. C.)............................... 9* 218 LVI. Comparison of Average Composition of Muscovites from FranklinSylva, N. C., and Petaca, N. Mex., Districts.................. 222 LVII. Two Muscovites Analyzed by Liashchenko (1940)................... 223 LVIII. Micas from the Schuttenhofen Pegmatite Analyzed by Scharizer (1887B)............................ 223 LIX. Two Micas from Sakihama, Japan, Analyzed by Shibata (1952A)...... 224 LX. Average Contents of Various Oxides in Micas..................... 226 LXI. Chemical Evolution of Muscovite.................................. 227 xi

LIST OF FIGURES Page lo Muscovite structure after Grim and Bradley (1951)...o... s*. & 8 2. Muscovite showing some diffuse scattering along 02 reciprocal lattice line. 0-level a-axis. Compare with Figure 5.......,. 48 3. Stereographic projection of established forms for muscovite after Peacock and Ferguson (1943)e.........o...o............. 54 4. Crystal drawing of muscovite forms found by Peacock and Ferguson (1943)...........o. ooo..o.oo..0o0,0ooo,... 54 5. O-level a-axis Weissenberg photograph of normal muscovite... o.,< 86 6. 0-level a-axis Weissenberg photograph of lithian muscovite,.....o 87 7. X-ray powder photographs of polymorphs in muscovite-lepidolite series............o..eoooooo..oooooooooo.....................0. 140 8. Plot of the Li20 content of 35 micas in muscovite-lepidolite series against the various polymorphs Oo o oo oo....o.....o. 8,.A 144 9. Idealized representation of relationship between Li20 content and polymorphism in the muscovite-lepidolite series.o........, 144 100 O-level pseudo a-axis Weissenberg photograph obtained from 3layer monoclinic lepidolite............................. 148 11o O-level b-axis Weissenberg photograph obtained from 3-layer monoclinic lepidolite............. oo o o oo o o............o o o o o.o o o 149 12. 1-level a-axis Weissenberg photograph obtained from 3-layer monoclinic lepidolite.....o o o o o e...... o o o o o.............o o o o o..............o e oo 149 13. O-level a-axis Weissenberg photograph of zinnwaldite with 2-layer monoclinic structure,......e..ee..., oeo o o o..,... oo 203 x 1i

ABSTRACT The micas, a group of sheet-structure alumino-silicates, are composed mainly of K,Mg,Fe2,Fe,Mn and Li with OH and F but also contain varying small amounts of Na,Ba,Rb,V,Cr, and Ti as well as minute amounts of a large variety of trace elements. The main species are muscovite, lepidolite, phlogopite, biotite, zinnwaldite, paragonite, roscoelite, and taeniolite. Most micas are monoclinic; some are hexagonal; a few may be triclinic. In muscovite the 2-layer form is essentially invariant. Lepidolites crystallize in 1-layer, 3-layer and 6-layer forms which can be correlated with the Li content. Phlogopites, biotites, and zinnwaldites also show considerable polymorphic variation but correlations with composition or occurrence are not yet possible. The micas of commercial significance are muscovite and phlogopite. Economically significant deposits of sheet-size, high-quality muscovite are confined to pegmatites, principally in wall zones and core-margin units. Muscovites from different zones in individual pegmatites are characterized by distinct differences in color, structural defects, and composition. Similar variations occur among the micas of different districts. The peratitic micas show systematic differences in composition, mainly in Fe3,Fe,Mg,Li, and F, which reflect the stage of Differentiation of the pegmatite magma at the time each type crystallized. xiii

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN PART I. GENERAL INFORMATION I. INTRODUCTION A. General Statement The micas constitute a closely related group of minerals with both common and rare representatives. This report gives the results of research on this group of minerals undertaken mainly as Project M978, Natural Mica Studies, of the University of Michigan Engineering Research Institute, sporsored by the United States Army Signal Corps. The purpose of the project was "To conduct investigations on physical properties, crystal structure, and composition of natural micas as related to geological occurrences and genesis." Thus we have attempted to learn as much as possible concerning the natural history of micas. A study of the mica group is of significance for two reasons: (1) Certain members of the group are widely distributed rock-forming minerals and therefore a study of these species leads not only to information regarding the conditions of their own formation, but also helps in establishing the environmental conditions under which the enclosing rocks were developed. (2) Several of the micas are of crucial and strategic significance, either as raw materials themselves or as a source for rare metals. We have attempted to concentrate our work mainly on those micas that are of some economic significance; these include muscovite, phlogopite, and lepidolite. But because of the difficulty of defining the limits of the various species, and also because of the structural and chemical relationships between the economically significant micas and others of lesser or no commercial importance, some of our work has necessarily included micas of no direct economic value. Members of the mica group are widespread in igneous rocks and mineral deposits of magmatic derivation; for example, muscovite in certain granites and particularly in granitic pegmatites; phlogopite in certain peridotites; biotite in gabbros, norites, diorites, and granites as well in many granitic pegmatites and other types of pegmatites; and lepidolite and 1

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN zinnwaldite, mainly in certain rare pegmatites and some high-temperature veins. Muscovite and its fine-grained variety, sericite, are widespread essential constituents and accessory minerals of hydrothermal veins and replacement deposits. Among the metamorphic rocks, biotite and/or muscovite are common and widely distributed in many types of slates, phyllites, schists, and gneisses, and phlogopite is a common constituent of some marbles. In sedimentary rocks, fine-grained detrital muscovite may be an important mineral in clastic rocks and authigenic muscovite may also become a significant constituent in some fine-grained elastics. We have not attempted to study the micaceous minerals that lie between the true micas and the clay-minerals group, namely, members of the illite group of minerals of which the most widespread representative is hydromuscovite or hydromica. The micas are also of scientific significance in that they are the most outstanding representatives of a type of silicate structure, namely, the sheet structure or disilicate type, in which the silica tetrahedra all are placed in one plane with each tetrahedron being joined to others by the three oxygen atoms that lie within the common plane. A continuous extension of such a linkage provides a hexagonal network within the plane. Other mineral groups that are characterized by this basic sheet-type structure include the chlorites, brittle micas, and vermiculiteso By studying representatives of the mica group it is possible to obtain more information on the fundamental properties of silicates showing this type of linkage. The micas are of scientific significance also because of their highly diversified composition, being hydrous or fluosilicates of alkali metals, principally potassium, and of divalent as well as trivalent metals. Conse — quently, a great many isomorphous substitutions are possible within the group. From the economic standpoint, muscovite and phlogopite represent substances that require an essentially unique type of mining and recovery, for they are desirable riot because of their composition, but mainly because of their unusual combination of physical properties and because of the size of-.the crystals in which they occur, Consequently, the mining of muscovite and phlogopite demands that the crystals as they occur in the deposit be removed without damage or without diminution of grain size. In this respect the micas are similar to quartz crystals in the requirements of their mining. The unique combination of properties possessed by muscovite and phlogopite which makes them economically desirable is: (1) their perfect basal cleavage, (2) their flexibility and elasticity, (5) their extremely low electrical conductivity and high dielectric property, (4) their very low heat conductivity, and (5) their occurrence in large, relatively perfect, single crystals. Lepidolite, the other mica of commercial value, is not mined because of its properties as a mica, but because it contains the element lithium which is extracted from it. Thus it is unimportant whether it occurs in large single crystals or relatively free of mineralogical impurities. 2

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN The importance of the mica studies is therefore two-fold; scientific, leading to information on mineral and rock genesis, silicate structures, and isomorphism; and economic, leading to data on occurrences, recovery, and effective utilization. B, History and Personnel of Project The investigations of Heinrich on micas began in 1941 at the suggestion of the late Professor Harry Berman of the Department of Mineralogy and Petrography of Harvard University. It was originally intended to attempt a study of the micas for a doctoral dissertation, but due to the death of Professor Berman and various other factors, this was abandoned. However, this preliminary work led directly to the compilation of a relatively complete collection of analyses of micas and a bibliography of articles relating to all aspects of mica research. From this work, one paper (Heinrich, 1946) resulted. During the period 1942 to 1947 Heinrich, as a member of the United States Geological Survey, was assigned during much of this time to the investigation and mapping of pegmatite mineral deposits, particularly mica pegmatites, and participated in such investigations in the southeastern mica districts of North Carolina, Georgia and Alabama as well as those of Idaho~ Colorado and New Mexico. Subsequently, he has had opportunity to study mica deposits in New England, Montana and other western states, in Canada, and in various European countries, particularly those of Scandinavia and Germany. Most of the mica specimens which have been studied as part of this investigation have been personally collected by Heinrich and were carefully labeled and studied as to their detailed geological occurrence within their deposits. One group of Colorado specimens was collected by Levinson and the remaining studied specimens are analyzed micas which have been obtained from numerous investigators both within the United States and abroad. In all cases, however, we have attempted to confine our studies to specimens whose detailed paragenesis either was known to us from our own field work or whose geological occurrence was a matter of record. General mica investigations were resumed in 1950 by Heinrich and Levinson in an attempt to learn more regarding the variation in optical and physical properties of muscovites within individual pegmatites and individual pegmatite districts. Attention was soon directed toward the widespread polymorphism existing in the mica group, and it became clear that in order to learn more regarding the genesis of pegmatitic muscovites, basic information on structural and compositional variations was required. In September, 1951, a large part of the investigation was transferred to Project M978 of the Engineering: Research Institute of the University of Michigan sponsored and supported by the U. S. Army Signal Corps, without whose assistance the wide scope of our work would have been impossible. 5

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN At the Mineralogical Laboratory of the University of Michigan, many techniques were utilized in the quest for new data on natural micas. Foremost of these were X-ray techniques, by means of which the structures of over 1000 specimens of mica were identified. The refractive indices of about 500 specimens, particularly muscovites, were determined by means of the Abbe" refractometer and immersion liquids, and 2V was measured by means of the Fuess axial-angle apparatus and the Mallard method. Extensive surveys of the literature were undertaken in connection with such problems as the study of overgrowths, zoning, paragenesis, and trace elements, in an attempt to correlate and interpret the older data on these subjects with information secured from our measurements and from the large number of new chemical analyses. A number of apparently significant Russian papers were translated in their entirety. The project during the two years of its sponsorship by the U. S. Army Signal Corps has been under the administrative supervision of Dr. S. Benedict ILevin, Deputy Chief, Chemical-Physics Branch, Squier Signal Laboratory. The project as a whole has been guided and supervised by Heinrich. Levinson has personally carried out nearly all of the X-ray work. The other two writers, Levandowski and Hewitt, have been employed as general assistants in varied aspects of the research work, but have also contributed specifically to the writing of certain sections of this report. C. Acknowledgments We are extremely grateful for the assistance rendered by many indi — viduals and organizations. Technical assistance and advice were given by Dr. S. B. Levin, Dr. H. Kedesdy (Squier Signal Laboratory); Mr. C. E. Harvey (Amer. Spectrographic Laboratories); Dr. R. A. Hatch, Dr. J. A. Kohn (U. S. Bur. of Mines); Prof. L. S, Ramsdell, Prof. R. M. Denning, Prof. C. B. Slawson, Prof. G. B. B. M. Sutherland (Univ. of Michigan); Dr. H. S. Yoder (Geophysical Lab.); Dr. W. T. Schaller, Dr. M. Fleischer (U. S. Geological Survey); and Prof. S. Goldich (Univ. of Minnesota). Through the courtesy of the following investigators we have been fortunate in obtaining samples of about 200 different micas, of which approximately 150 have been analyzed and described in the literature: (1) Dr. J. M. Axelrod (U.S.G.S., Washington) (2) Dr. F. A. Bannister (Brit. Mus. Nat. History, London) (3) Prof. L. G. Berry (Queens Univ., Kingston, Ontario) (4) Mr. J. E. Bever (Univ. of Michigan) (5) Dr. Harald Bjorlykke (Norwegian Geol. Survey) (6) Dr. H. V. Ellsworth (Geol. Survey of Canada, Ottawa) (7) Mr. R. C. Erd (U.S.G.S., Washington) 4

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN (8) Prof. H. W. Fairbairn (Mass. Inst. Techn.) (9) Prof. R. R. Franco (Univ. of Sao Paulo, Brazil) (10) Prof. C. Frondel (Harvard Univ.) (II) Prof. E. N. Goddard (Univ. of Michigan) (12) Dr. C. H. Greene (Corning Glass Works, Corning, N.Y.) (13) Mr. W. R. Griffitts (U.S.G.S., Washington) (14) Dr. E. Grip (Boliden, Sweden) (15) Dr. A. F. Hallimond (London, England) (16) Dr. S. B. Hendricks (U. S. Dept. of Agriculture, Washington) (17) Dr. P. Hidnert (Nat. Bur. of Standards, Washington) (18) Prof. W. F. Hunt (Univ. of Michigan) (19) Prof. C. 0. Hutton (Stanford Univ., California) (20) Mr. R. Jacobson (Geol. Surv. of Nigeria, Kaduna Junction) (21) Prof. J. Jakob (Tech. Univ. of Zurich, Switzerland) (22) Dr. Y. Kawano (Geol. Surv. of Japan) (23) Dr. A. Knopf (Yale University, New Haven, Conn.) (24) Prof. L. Lokka (Helsinki, Finland) (25) Mr. A. M. Macgregor (Geol. Survey of Sourthern Rhodesia) (26) Prof. C. Mauguin (Univ. of Paris, France) (27) Dr. V. B. Meen (Royal Ontario Museum, Toronto) (28) Dr. and Mrs. A. Miyashiro (Geol. Surv. of Japan, Tokyo) (29) Prof. J. Murdoch (Univ. of California, Los Angeles) (30) Prof. T. Noda (Nagoya Univ., Japan)(31) Prof. E. W. Nuffield (Univ. of Toronto, Canada) (32) Dr. O. H. Odman (Royal Inst. of Technol., Stockholm) (33) Prof. I. Oftedal (Univ. of Oslo, Norway) (34) Prof. G. P. Pagliani (Univ. of Milan, Italy) (35) Prof. R. L. Parker (Techn. Univ. of Zurich, Switzerland) (36) Prof. G. Pehrman (Abo Academy, Abo, Finland) (37) Prof. R. Pieruccini (Univ. of Florence, Italy) (38) Dr. F. H. Pough (Amer. Mus. Nat. Hist., New York) (39) Prof. R. T. Prider (Univ. of Western Australia, Perth) (40) Prof. Percy Quensel (Univ. of Stockholm) (41) Mr. A. H. Roberson (U. S. Bur. of Mines, Albany, Oregon) (42) Dr. H. P. Rowledge (Govt. Chem. Laboratories, Perth, West. Australia) (43) Dr. A. Schiener (Mus. Nat. Hist., Vienna, Austria). (44) Dr. F. T. Seeyle (Dominion Laboratories, Wellington, New Zealand). (45) Dr. K. Sugiura (Toyko Inst. of Technology, Japan). (46) Dr. G. Switzer (Smithsonian Institution, Washington) (47) Prof. S. Tyler (Univ. of Wisconsin, Madison, Wis.j (48) Dr. S. van Biljon (Heidelberg, Transvaal, South Africa) (49) Dr. F. Wickman (Royal Mus. of Stockholm, Sweden) (50) Prof. H. Winchell (Yale Univ., New Haven, Conn.) (51) Dr. H. Yamada (Tokyo Inst. of Tech., Japan).. ---------. --------------------— 5

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Mrs. E. G. Smith provided Russian translations, and Miss M. J. Abels and J. I. Smith, French translations. S. J. Lefond, A. A. Giardini, -and A. Geyer assisted in the laboratory, while Mrs. S. Roof and Miss J. L. Everett furnished secretarial aid. The Faculty Research Fund of the Horace H. Rackham School of Graduate Studies, University of Michigan, furnished the funds for five new chemical analyses. 6

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN II. CRYSTAL CHEMISTRY OF MICAS Mauguin (1927, 1928A, 1928B) was the first to apply X-ray analysis to the study of micaso He determined their unit cells and symmetry and discussed the types of isomorphous replacement. Knowing the cell dimensions, the density and composition of micas, and the weight of each atom, he calculated the average number of atoms of each kind in the unit cell rOne of the main points noted was that the total number of oxygen and fluorine atoms in the structural unit is always 12. It is upon this value that modern formulas of the micas are based, Pauling (1930) advanced the general concept of the structure of mica and other layered silicates by proposing on the basis of Alis coordination theory, an outline of the general structural scheme for these mineralso Paulingls (1930) work gives, in particular, the sequence of atomic planes parallel to the cleavage plane. The first complete analysis of a mica structure on this basis was made independently by Jackson and West (1930, 1933)o Their X-ray work on muscovite showed conclusively that the ideas proposed by Pauling (1930) were correct. The basis of the mica structure, which has been described in detail by Bragg (1937), Hendricks and Jefferson (1939), and others, is the silica tetrahedron (Si04) arranged in sheets to form a hexagonal network (Fig. l). Silicon has a coordination number of four, In the tetrahedron, aluminum may replace one-fourth of the silicon. Inasmuch as the silicon-oxygen sheet, by sharing oxygen atoms of adjacent tetrahedra, has the composition (Si4010); the substitution referred to may result in the silicon-oxygen. sheet having the formula (A1Si3010)o Two of these sheets are placed together with the vertices of the tetrahedra pointing toward each other. In muscovite these sheets are held together by aluminum atoms which have octahedral coordination with respect to the neighboring hydroxyl groups and oxygen atoms. The hydroxyl groups are situated at the centers of the plane hexagons formed by the oxygen atoms of the silica tetrahedra, The octrahedral groups may also be built around magnesium as in phlogopite, lithium as in lepidolite, and so forth. This constitutes one firmly bound double-sheet. The structure of the micas is completed by joining the above-mentioned double-sheets with potassium atoms which have twelve-fold coordination with respect to oxygen atoms. The perfect cleavage of the micas lies between these double-sheetso With the possible exception of titanium, the position occupied by the many elements reported in all good analyses is no longer i.n doubt, Berman (1937) indicates the general formula for the mica group as~ ----------------------- 7 —----------

-— __+215)l' 1 I A1:K......-4(.;(i -3Si4-, ]AI3+ 7 -3Si,6 02— 6 ~ " L _ —4. O0.201 1 Fig. 1.Muscovite -strutr — 3S-, r1 A13+ Fig. 1. Muscovie s —-t r 2K Fig. 1. Muscovite structure after Grim and Bradley (1951).

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN W(X,Y)2_3z40lo( (OHF)2 in which W = K predominantly; Na, Ba, Ca, subordinate to rare; X = Mg, Fe2, Mn2 Li; Y = Al, Fe3, Ti subordinate, Cr, Mn3, V rarely; Z = Si-Al 5:3 to 7:1 Atoms represented by W have twelve-fold coordination, those represented by X and Y have octahedral coordination, and those represented by Z have tetrahedral coordination. Mauguin (1927, 1928A, 1928B) noted in his original X-ray studies that the c-axis of biotite was apparently only half as long as that of muscovite and he was thus the first to encounter polymorphism in the micas. It remained for Hendricks and Jefferson (1939) to give a full account of the scope and magnitude of polymorphism in this complex group. In all, among 100 different specimens, they discovered seven polymorphic variations which embrace three different crystal systems. They described the seven different structures (hexagonal, monoclinic, and triclinic) in detail and showed how they might be derived from the single-layer form by the application of various combinations of symmetry operations. Hendricks and Jefferson (1939) pointed out that muscovite crystallized in only the two-layer muscovite structure, but Axelrod and Grimaldi (1949) have reported a three-layer monoclinic muscovite (see page 47). Levinson (1953), as a result of work carried out on this project, has described a new variation of the muscovite structure for which the term "lithian muscovite" has been proposed. These are the only X-ray studies on micas since the classic contribution of Hendricks and Jefferson (1939). 9

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN III. X-RAY TECHNIQUES FOR IDENTIFICATION OF POLYMORPHS It has been necessary to do a considerable amount of X-ray work, consisting mainly of identifying the structures of as many micas as possible. Whenever possible, Weissenberg photographs were taken about the O-level a-axis, as this type of photograph is unique for the different polymorphs. In many cases, however, upper level photographs were taken as further verification of the structural interpretation. In determining the orientation of the flakes used in the single crystal work, an optical procedure was used whenever practicable. Lines corresponding to the desired crystallographic axes were scratched upon the cleavage flakes with a razor. The exact positions of these axes were determined from interference figures and from extinction positions in those micas which have observable birefringence on cleavage planes. For all practical purposes, the crystallographic axes may be assumed to correspond with the extinction postion. However, many micas, particularly biotites and phlogopites, are uniaxial or have small 2V. In this case the orientation had to be determined from Laue photographs or by means of percussion figures. In several instances crystal development was poor, crystals were minute, or extinctions were poor. This made single-crystal methods difficult if not impossibleand left powder X-ray methods as the only study technique. Unfiltered copper radiation was used in most single-crystal work and a nickel filter (with copper radiation) was used in powder studies. Iron radiation was used whenever powder photographs of biotites were taken. Only a large diameter (114.6 mm) powder camera is suitable for the mica studies. Polymorphs of muscovite and lepidolite can be distinguished by means of X-ray powder photographs but those of biotite and phlogopite cannot be differentiated by this method. 10

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN IV. NOMENCLATURE In every complex mineral group there exists a large number of varietal names based on major or minor chemical variations, optical distinctions, grain size, occurrences, localities, and misidentification. The mica group is no exception for it is burdened with a surplus of trivial names. The list that follows defines the varietal names that have been employedin the series. Most of these are listed in Hey (1950), Chemical Index of Minerals, but the following references also were consulted in constructing the complete synonymy. 1. Dana (1892), A System of Mineralogy. 2. Chester (1896), A Dictionary of the Names of Minerals. 3. Winchell and Winchell (1951), Elements of Optical Mineralogy, Part II, Description of Minerals. The following scheme is employed to differentiate the various categories: Synonyms and varieties: lower case (e.g., adamsite) End-members: first letter capitalized (e.g., Ferrimuscovite) Pseudomorphs: capital delta preceding name (e.g,, Aachlusite Doubtful species: underlined (e.g., anhydromuscovite) Valid species, chem. varieties: capitals (e.g., MUSCOVITE) Terms which are useful or sufficiently well defined, in the opinion of the Writerp, and should therefore be retained: asterisk preceding name (e.g., *MUSCOVITE). Aachlusite: a green alteration product of topaz, resembling steatite, but near soda mica in composition; pseudomorphous after topaz. adamsite: syn. of muscovite. agalmatolite: syn. of pinite, pyrophyllite or talc. alurgite: var. of muscovite near K2(Mg,Al)4_5(Al,Si)8020(OH)4J also contains about 1% Mn oxides. 11

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN ammochrysos: syn. of muscovite. amphilogite: syn. of didymite. anhydrobiotite: artificially dehydrated biotite, nature unknown. anhydromuscovite artificially dehydrated muscovite, nature uncertain. annite: var, of lepidomelane; syn, of hydroxyl-anniteo anomite: varo of biotite with optic plane normal to b-axis (see meroxene). anthrophyllite: "A mica (?)" Hey (1950). Aaspasiolite: aluminosilicate of Mg, an altered cordierite. aspidolite: near (NaK)2(Mg,Fe+2) l2Si 00(OH)4 with Na:K 53 and Mg:Fe2e 5; near phlogopite. avalite: near K(Al,Cr)4(Al,Si)8020(0H);4 probably a mixture (Chester, 1896). baddeckite: a mixture of hematite and clay, originally considered to be iron mica related to muscovite (Schaller and Henderson, 1926)o barium muscovite: syno of oellacherite. barium-phlogopite: var. of phlogopite with 1% BaO, barytbiotite: var. of phlogopite; 2 (KBa)2(Mg,Al)4_6(A, Si)8020(0H)4] containing 6.84% BaO and also some Fe. Analysis in Dana (192) shows it to be a phlogopite rather than biotite (Hey, 1950, p. 160). basonite: near a hydrobiotite or a vermiculite, near K1/2(Mg,Fe3,Fe2Al) (SiAl)802 (OH)41-1/2H20 with Mg and Fe o2 and Sie 5-l/2. bauerite: SiO2 pseudomorphs after mica; the nature of the material is unknown. (Hey, 1950) biaxial mica:~ syno of muscovite. bildstein: syn. of agalmatolite. *BIOTITE: K2(Fe2,Mg) (Fe,Al,Ti)O (Si6 Al )0 (OHF)4 2 6-4 0-2 6-5 2-3 20-22 4-2. Abonsdorffite: aluminosilicate of Fe2 and Mg, an altered cordierite, pinite group. 12

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN brammallite: var. of illite, aluminosilicate of Na and K; syn.is sodium illite. buldymite: aluminosilicate of Mg,Fe2,Fe3, and K; probably a hydrobiotite. caesium-biotite: var. of biotite with as much as 3% Cs20 and 1-1/2% Rb20 (Hess and Fahay, 1932). calciobiotite: var. of biotite containing as much as 14% CaO. Acataspilite: aluminosilicate of Mg,Ca,Na, and K, pseudomorphous after cordierite. cat gold: syn. of muscovite; an early popular name for gold-colored mica. catlinite: a mixture of sericite, pyrophyllite, etc. cat silver: syn.of muscovite. chrome mica: (1) syn. of fuchsite (!Daha, 1892, p. 614); (2) syn. of fuchsite and mariposite (Whitmore et al., 1946). chromglimmer: syn. of chrome-mica and fuchsite. chromglimmer (of Schafhautl): in part var. of biotite, near K2(Mg,Fe,Cr,Al)4(AlSi)8020(OH)4; the rest fuchsite (Hey 1950). *CHROMIAN MUSCOVlTE: name propose'd for fuchsite since fuchsite has the typical muscovite structure (Whitmore, et al., 1946). CHROMIAN PHENGITE: name proposed for mariposite as it has a high silica content and a small (less than 1%) Cr203 content. (Whitmore, et al., 1946), chromochre: appears to be similar to niscovite but has more Cr203 and has a very small optic angle; it may be a chromian phengite (Winchell and Winchell, 1951). colomite: syn. of roscoelite. common mica: syn. of muscovite. Acordierite-pinite: pinitic pseudomorph after cordierite. cryophyllite: var. of zinnwaldite often with some deficiency in (Li,A1,Fe) group; typically Li %2-1/2, (Fe2,Fe3) 1, and Si - 7. Also the name has been used for 2 hypothetical end-members; K2Li2FeA12Si7O18 (OH)4 and K2Li2FeAl2Si6016(OH)4 (Hey, 1950). cymatolite: muscovite and albite mixture, perhaps pseudomorphous after spodumene. 13

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN damourite: var. of muscovite; an altered muscovite, usually contains more water or more easily removed water (Winchell and Winchell, 1951). Folia less elastic; includes most hydro-mica in general; does not necessarily contain more water; they may, however, give it off more easily (Dana, 1922). didrimite: syn. of didymite. didymite: syn. of muscovite. Adysyntribite: syn. are dysintribite, dyssintribite, dyssyntribit; aluminosilicate of K and Na; pseudomorphous substance of pinite group. Eastonite (of Winchell): theoretical end member, K2Mg5A14Si520p(OH)4; some natural micas apparantly approach this composition closely. epileucite:' pseudomorphs of orthoclase and muscovite after leucite. epi-sericite: syn. of sericite; sericite formed in epi (upper) zone of metamorphism. euchlorite (of Shephard): var. of biotite. eukapt alinosilicat of g and K; a mica or chlorite, not well defined. euphyllite: near (Na,K)A13Si010 (OH)2; may be a mica intermediate between muscovite and paragonite, or a brittle mica, or perhaps a mixture. Afahlunite: aluminosilicate of Mg, Fe, and K near (Mg,Fe2) A12Si3012H20;O pinite group. ferribiotite: syn. of lepidomelane. Ferri-muscovite: hypothetical end-member used by Winchell (1951) (K2Fe32A12 (OH)4Si6A1202o); also by Wahl (1925)(K Fe33Si3010(OH)2). ferrititanbiotite: syn. of ferriwotanite. ferriwotanite: var. of biotite high in Fe3 and Ti (see wodanite). ferro-ferri-muscovite: syn. of monrepite. Ferrophengite: hypothetical end-member used by Winchell and Winchell (1951) (K2Fe2A1 (OH )4Si7A102) o Fluor-annite: var. of lepidomelane; the mica end-member KFe2Al1Si5O10F2. -----------------— ^ -------------------

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN Fluor-biotite: var. of biotite; the mica end-member K(MgFe2)AlSi3O10F2 Fluor-lepidomelane: var. of lepidomelane; end-member K(Mg,Fe2)3(Al,Fe5) Si010F2~ Fluor-meroxene: var. of biotite; the end-member KMg(Al,Fe3)Si3010F2. Fluor-phlogopite: var. of phlogopite; the fluorine end-member KMg3AlSi3010F2. Also the most abundant synthetic phlogopite. Fluor-siderophyllite: var. of biQtite; the end-member KFe2 (AlFe )Si310F 2. frauenglas: syn. of muscovite. fuchsite: var. of muscovite with as much as 4.81% Cr20O3 X-ray studies by Whitmore, Berry, and Hawley (1946) have shown that fuchsite has the muscovite structure. They propose that chromian muscovite be adopted in place of fuchstte which has no structural significance. This proposal is entirely justified. It follows, therefore, that if mariposite is also shown to have structures characteristic of phengite, the term chromian phengite or hexagonal chromian phengite would be appropriate (owing to a high silica content). gaebhardite: syn. of fuchsite. Agieseckite: aluminosilicate of Mg and K; sometimes appreciable FeO; a pinite type pseudomorph. gigantolite:- a mixture of muscovite and biotite; pinite group. gilbertite: var. of muscovite. glimmer: syn. of mica. goeschwitzite: syn. of illite. Agongylite: aluminosilicate of Mg,Fe2. and K; a pinitic pseudomorph. grundite: syn. of illite; trade name for illite clayo gumbellite: var. of hydromuscovite (Hey, 1950); a hydrous silicate of aluminum, near pyrophyllite (Chester, 1896). hallerite: a lithium bearing mica (0.5 to 1.5% Li), but it is not clear whether it should be classed as a variety of paragonite or of muscovite. 15

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN haughtonite: var. of biotite rich in Fe2 and Fe3. helvetan: aluminosilicate of Fe2,Mg,Ca and alkalis; probably an impure muscovite *heptaphyllite: group name for those micas which contain 7 other atoms to 12 (O,OHF) as distinguished from octophyllite micas which contain 8 atoms to 12 (O^OHF); originally applied to muscovite and lepidolite but now applies to muscovite only (and paragonite). The term heptaphyllite is structurally correct, and is thus considered valuable. heterophyllite: var. of biotite; a mica between siderophyllite and annite; somewhat lower in Fe2 than siderophyllite; the name seems of doubtful value o hexagonal mica: syn. of biotite. hydrobiotite: a hydrated biotite low in K, Mg, etc. and OH in place of the 0. Also classed as a vermiculite. hydromica: "Used as a term to designate a micaceous clay mineral of common occurrence resembling sericite but having weaker double refraction. Galpin (1912) indicates that kaolinite alters through metamorphism into hydromica, thence to muscovite... The group of hydrous micas, illite, hydromuscovite or glimmerton has been studied with particular care by Grim, Bray and Bradley (1937) who proposed the name illite..0. The material has less K20 and more water than muscovite," (Kerr and Hamilton, 1949 p. 33) *hydromuscovite: (1) var. of muscovite higher in OH, lower K, or K and Al than muscovite (Hey, 1950) (2) general name for various hydromicas derived from muscovite (Chester, 1896). (3) This mineral seems to lie between muscovite and illite both as regards potassium content and perfection of crystallinity (Brammall, Leach and Bannister, 1937). hydrophlogopite: a var. of phlogopite high in H20. hydrous mica: syn of hydromica. Hydroxyl-annite: var. of lepidomelane; the end-member K2Fe26(A2Si6)020(0H) 4 with little Mg (Grigoriev, 1935). Also a syn. of annite. (Winehell and Winchell, 1951) -- ------- 15 —6

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Hydroxyl-biotite: syn. of normal biotite so named to distinguish it from f luor-biotite. Hydroxyl-lepidomelane: syn. of lepidomelane, in particular the F-free end-member. Hydroxyl-meroxene: var. of biotite; the hydroxyl end-member to be distinguished from fluor-meroxene. Hydroxyl-phlogopite: syn. of normal phlogopite so named to distinguish it from fluor-phlogopite. Hydroxyl-siderophyllite: syn. of normal siderophyllite so named to distinguish between it and fluorsiderophyllite. Ahygrophyllite: aluminosilicate of Fe2,Mg,Ca and alkalis; derived in part, at least, from feldspars; pinite group. iberite: (of Svanberg, 1844) syn. of gigantolite. *illite: "It is not proposed as a specific mineral name, but as a general term for clay mineral constituents of argillaceous sediment belonging to the mica group." (Grim, Bray and Bradley, 1937, p. 816). iron mica: syn. of lepidomelane and biotite. irvingite: lepidolite near (Na,K) (LiAAl)5(Si,A1) 8(,OH,F)24 with Lil 2-1/2 and (0H,F)w3. isinglass: syn. of mica, particularly muscovite and phlogopite. ivigtite: aluminosilicate of Fe and Na(?); analysis imperfect (Hey, 1950); may be near gilbertite (Chester, 1896). kaliglimmer: syn. of muscovite. Akillinite: aluminosilicate of K; pseudomorphous after spodumene; probably identical with muscovite; pinite group. kryptotile: A13(Si04)3H3, extreme hydrogen end of muscovite series (Clarke, 1914). lardite: syn. of steatite and of agalmatolite; pinite group. *LEPIDOLITE: K2 ( Li,A1 ) 5-6 (S i 6_7,A 12_1)020 - 21(FOH )4 17

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN lepidomelane a term commonly employed for iron-rich biotite. The term has been used for biotites rich in Fe3, those rich in Fe2, and those with relatively large amounts of both Fe2 and Fe3 (Heinrich, 1946). lepidomorphite: (1) var. of phengite (Hey, 1950); (2) a fine, scaly mica; the result of the alteration of oligoclase (Chester, 1896). leucophyllite: a high silica var. of muscovite; similar to phengite; original analysis possibly on impure materialo leverrierite: may be the same as kryptotile but of different genesis (Clarke, 1914)o Aliebenerite: aluminosilicate of alkalis of Fe and Ca; a pinite group pseudomorph o lilalite: syno of lepidolite. lilalitho syn. of lepidoliteo lithia mica: syno of lepidoliteo lithioneisenglimmer: syno of zinnwalditeo lithionglimmer: syno of lepidolite. lithionit: syno of lepidolite. lithionite: syn, of lepidoliteo lithionitesilicat: syn. of lepidolite. Lithium muscovite: a hypothetical mica end-member: (1) Stevens (1938), K4Li6AlSi12040 (OH,F ) (2) Berggren (1941), K4Li6All0Si12041F50H. lithium (Li)-phengite: contains 3.02% Li20, 47.44% SiO2, 1.41% Mg0 and 1087% total Fe. These values do not permit it to be classed as a phengite. Perhaps somewhere between muscovite and lepidolite; described by Shibata (1952A and 1952B). lythrodes: an altered nepheline; pinite group. macrolepidolite: var. of lepidolite; daistinguished from microlepidolite by larger 2V, 18

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN magnesia mica: syn. of phlogopite. mahadevite: a mica supposedly between muscovite and phlogopite in composition; near (K,Na)o 97(Al,Fe,Mg) 66(Si,A A)4(OOHj; should be reanalyzed. manandonite: a basic boro-silicate of lithium and aluminum closest to lepidolite in composition LiAl114B4Si6029(OH)24 (?); should be studied further; possibly not a mica, manganese mica: syn. of manganophyllite. manganese muscovite: syn. of mangan-muscovite. mangan-muscovite: var. of muscovite with about 2% MnO. manganophyll: syn. of manganophyllite. manganophyllite (of Igelstrom, 1872): aa Mn- var. of phlogopite; K MgF eFe2 J o 3 Mn3 ))(Si6Al2)020(OH)4; also considered var. of biotite (Hey, 19g07 Manganophyllite (of Yoshimura, 1959): hypothetical mica end-member K2Mn5A14 i5020 (oH)4' margarodit: syn. of margarodite. margarodite: syn. of damourite. marienglas: syn. of muscovite. mariposite: var. of phengite (Whitmore, et al., 1946) with high silica and as much as 1% Cr203; see discussion of fuchsite. meroxene: var. of biotite which has optic plane parallel with b-axis (perpendicular to 010) as contrasted with anomite. Most biotite was considered of this type (Dana, 1892). metabiotite: syn. of bauerite. metasericite: var. of muscovite; a hydrous mica classed under damourite. micarel: mica pseudomorphous after scapolite; pinitic. micarelle: syn. of micarel. microlepidolite: var. of lepidolite distinguished from macrolepidolite by its small optic angle. 19

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN minguetite: near (KNa)0o8(Fe2 MgFe3 A1)5.2(SAl)8020(OH)4 with (FeMg) 53.2 and Si 6.8; belongs either to chlorite or mica groups; nature uncertain. monrepite: KFet1 Fe''Si 3010(0H)2; a ferro-ferrimuscovite; should be studied further. *MUSCOVTITE: [KAJ1(Si,Al)40 (OH)2] nacrite (of Thomson, 1836): syn. of muscovite. natronbiotite: var. of biotite containing appreciable Na. natronphlogopite: var. of phlogopite containing appreciable Na. oblique mica: syn. of muscovite. *octophyllite: group name for those micas which contain 8 other atoms to 12 (0,OH,F) atoms. Biotite, phlogopite and lepidolite generally classed as octophyllites (see heptaphyllite). odenite: syn. of odinite, odite, and oderite. var. of biotite (Hey, 1950) and var. of muscovite (Chester, 1896). oellacherite: muscovite containing barium. oncophyllite: muscovite pseudomorphous after feldspar. oncosine: (1) aluminosilicate of Mg and K; a pinite group pseudomorph; probably a mixture of muscovite, quartz, etc. (Hey, 1950 ); (2) a cryptocrystalline mica classed under damourite (Ford, 1932). onkophyllit: syn. of oncaphyllite. onkosin: syn. of oncosine. Aoosite: aluminosilicate of Fe2 and alkalis; a pinitic pseudomorph. pagodite: syn. of agalmatolite. ~PARAGONITE: (NaK) 2A14 (Si6A2 ) 20 (OH ) parophite: a pinitic rock. Paucilithionite: a hypothetical mica end-member (K2Li3A15Si6020F 41). 20

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN ~PHENGITE: var. of muscovite higher in silica than typical muscovite, and with Mg.Fe in place of some Al; K2(Mg,Fe,Al)4(SiAll)020(0H)4. *PHLOGOPITE: K2(MgFe2)6(Si6A2)020 (H)4 pholidolite: aluminosilicate of Mg, Fe2 and K; perhaps a fluorine-free, Al-poor phlogopite; nature uncertain. phyllite: (1) (of Thiebaut, 1925) syn. of pholidoide (group name for aluminous glauconites); (2) a general term for micas, chlorites, clays and vermiculites used by French writers; (3) (of Thomson, 1828) usually regarded as identical with ottrelite. Picrophengite: a hypothetical mica end-member K2MgAl3(Si7A 1)020(01H)4. *pinite: a group of pseudomorphs, mostly of mica, after cordierite, nepheline, or scapolite; Heinrich (1950, p. 183), after a study of cordierite in pegmatite concludes, "Cordierite that crystallized in the magmatic pegmatite stage is not in equilibrium during the hydrothermal stage and is altered to pinite, which consists chiefly of muscovite, chlorite, and biotite in varying proportions. This change requires chiefly the introduction of potassium and hydroxyl by pegmatite solutions." Apinitoid: aluminosilicate of Fe2 and alkalis; a pinitic pseudomorph. Apolyargite: aluminosilicate of Ca and K; a pinitic group pseudomorph. poly-irvingiteo a var. of irvingite high in Si. Polylithionite: a hypothetical mica end-member. polylithionite: a lithium bearing mica (lepdiolite) with high (as much as 9% Li20) lithium content. potash micat syn. of muscovite. Protolithionite: a hypothetical mica end-member (Kunitz, 1924); K2Fe AlSi3010 (OH)4. protolithionite (of Sandberger): var. of zinnwaldite. pseudobiotite: an altered biotite high in H20; probably a hydrobiotite. pterolite: a mixture of aegirine and lepidomelane. 21

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN pycnophyllite: syn. of pyknophyllite. pyknophyllite: syn. of sericite. Apyrargillite: aluminosilicate of Fe2,Mg, and alkalis; pinite group. Apyrrholite: aluminosilicate of Ca and K; a pinite group pseudomorph after anorthite. rabenglimmer: var. of zinnwaldite high in Fe3. rastolyte: a hydrated biotite similar to voigtite. rhombengliimmer: -syn. of biotite. rhombic mica: syn. of phlogopite. *ROSCOELITE: has been regarded as a vanadium muscovite with as much as approximately 20 = V205; K2(VAl)4(Si6Al2)0 2(0H)4; new data is presented in this report (p. 212 ) rosellan: syn. of rositeo rosellite: syn. of rosellan. Arosite: aluminosilicate of Ca,Mg, and K; a pinitic pseudomorph. rubellan: aluminosilicate of Mg,Fe, and K; probably an altered biotite. sandbergerite-syn. of oellacherite. sarospatakite: syn. of illite, scale stone: syn. of lepidolite. schernikite: pink, fibrous var. of muscovite occurring intergrown with lepidolite. schuppenstein: syn. of lepidolite. *sericite: (1) var. of muscovite; also includes much pinite (Hey, 1950); (2) fine scaly muscovite united in fibrous aggregates and characterized by its silky luster (Ford, 1932); (3) the name usually confined to white mica which is secondary, often the result of alteration of feldspar (it has been suggested that it contains less potash and more water) (Winchell and Winchell, 1951). Sericite as used by Winchell 22

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN and Winchell (1951) is so well established in the field of petrography that any attempt to dislodge or replace it will be futile. Gruner (1948, p. 684) notes, "The exceedingly finegrained micas, which go under such names as sericite, illite, and hydromuscovite, are in a class by themselves. They are defect structures principally with regard to missing alkali cations. What takes the place of these ions? Analyses show definitely that as K decreases H20 increases." shilkinite: described originally by Merkulova (1939) as a new mineral but shown by Tchirvinskii (1948) to be identical with sericite., I siderischer-fels-glimmer. syn. of lepidolite. 35 siderophyllite- var. of biotite high in Fe2 with little Fe. sodium-illite: syn. for brammallite. sterlingite: var. of damourite. taeniolite: syn. of tainiolite. *TAINIOLITE: K Mg4Li2Si8020(F)4; the only member of the mica group without essential aluminum. Classed under phlogopite by Dana (1926) and under lepidolite by Winchell and Winchell (1951). talcite: syn. of damourite. Aterenite: aluminosilicate of Mg and K; an altered scapolite; a pinitic pseudomorph. titanbiotite: syn. of wodanite. titanglimmer: original form of titanmica. titanmica: group name for titaniferous micas. titanobiotite: syn. of wodanite. triclasite: syn.of fahlunite (pinitic pseudomorph), uniaxial mica: syn. of biotite. vanadinglimmer: syn. of vanadium mica. vanadium mica: syn. of roscoelite. 23

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN verdite: impure fuchsite used as an ornamental stone. voigtite: near K(Fe2MgFe3 6SiA (H4-1/2 H2O with Fe2Mg:Fe3:2:1 and Si-i 5-1/2; probably a hydrobiotite. waddoite: a specimen labelled waddoite in the University of Michigan collections appears as typical biotite. Hoy (1950) notes that it is incompletely described. Awilsonite: aluminosilicate of Mg and K; an altered scapolite; a pinite. wodanite: a var. of biotite containing appreciable quantities of TiO2 (as much as 12.5%). wotanite: syn. of wodanite., *ZIiNWALDITE: K2(Fe212Li2 Ak) (Si6_7Al2_1)020(F52 2 There seems to be some confusion as to whether zinnwaldite is more closely related to biotite or lepidolite. kweiaxiger glimmer: syn. of muscovite. The preceding classification is in general agreement with Hey (1950); a few points of difference, however, should be discussed. Hey considers brammallite, for example, a valid species when actually it is no more than a sodium-bearing illite. His opinion on varietal names is expressed as follows: (p. XI): "In the matter of varietal names, the chemical prefixes of W. To Schaller (for example, Plumbian aragonite for Tarnowitzite) have not been adopted; certainly a great multiplication of varietal names based on small differences in chemical composition is to be deprecated, but it seems quite unnecessary to discard such well-known names as Freihergite, Pisanite, or Ceylonite; moreover, it is not really possible to draw a definite line between species and varieties.V Many objections to Hey's position on this matter have been raised and are clearly stated by Heinrich (1951., p. 634): "It is unfortunate that the compiler has seen fit to reject the Schaller System of adjectal modifiers which is rapidly obtaining general adoption by mineralogists in all parts of the world as a satisfactory scheme for freeing mineralogical nomenclature of its present unwieldy and largely meaningless agglomeration of varietal names. It is only by applying the -----------------— ^ ------------------

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN system of adjectal modifiers that mineralogists can remain consonant with a modern conception of minerals as constituting chemical series."' 25

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN V. CLASSIFICATION A. History of Mica Studies Hypotheses advancedprior to 1924 to explain the chemistry of the micas and the other silicates are interesting today only from a historical point of view but a brief resume of tm i u i g a i of them isuseful in giving an idea f the development of some modern theories. Silicates as a group had been recognized about 1780 by Bergmann. He included in this division minerals of the mica, amphibole, garnet, feldspar, and clay groups, as well as individual species. As the discovery of new silicate species continued, several theories were evolved to explain their complex chemistry. Dana's fifth edition of A System of Mineralogy (1868) presented one of the most complete and widely accepted presentations in English up to that time. Dana classified all silicates into two large groups —the anhydrous silicates and the hydrous silicates, each of which had its own subdivisions. The subdivision of the anhydrous silicates was as follows: 1. Bisilicates: oxygen ratio for the bases and silica 1:2 2. Unisilicates: oxygen ratio for the bases and silica 1:1 3. Subsilicates: oxygen ratio for the bases in silica 1: less than 1; mostly 1:2/3- but also 1:1/2 and 1:3/4 The following was the general subdivision of the hydrous silicates. 1. The general section of hydrous silicates. Includes all hydrous silicates except the zeolites and margarophyllites. a. Bisilicates. b. Unisilicates. c. Subsilicates. 2. Zeolite section. 3. Margarophyllite section. 1 ------ --------------— 26 -------------

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN The bases, with reference to the oyxgen ratio (defined by Danaj 1875, p. Xi as "the ratio between -the-number of atoms of oxygen in the different oxygen compounds present") for bases and silica, comprise such elements as potassium, sodium, lithium, magnesium, manganese, aluminum, etc. In relation to the element silicon, which was considered so strongly negative, the oxides of the above named elements were relatively basic. For all practical purposes, the bases were considered mutually replaceable, eight to ten often occurring in the same compound combine-d in either simple or intermediate ratios. Most of the micas were assumed to be unisilicates of the anhydrous silicate group, for the tersilicate subdivision, in which Dana placed them in his fourth edition (1854), was proved to have no existence. A few rare varieties such as margarodite, pinite, paragonite, oellacherite and euphyllite were placed in the margarophyllite section along with talc, serpentine and other hydrous micaceous minerals. Ratios other than for bases and silica formed the foundation for further subdivision within the anhydrous unisilicate group. Chief of these were the RO and R203 ratio, as follows: 1. Oxygen ratio: for bases and silica, 1l; Phlogopite - oxygen ratio for RO, R203, between 2:1 and 5:3; Biotite - oxygen ratio for RO, R203O about 1:1; Lepidomelane -oxygen ratio for RO, R203, about 1:3; Annite - oxygen ratio for RO, R203, 1:2. 2. Oxygen ratio for bases and silica, 1:1-1/4 to 1:2; Muscovite - oxygen ratio for RO, R203, 1:6 to 1:12 and for RO + R203, SiO2, mostly 1 1-1/4; Lepidolite - oxygen ratio for RO + R203, SiO2 mostly 1 1-1/2; Cryophyllite - oxygen ratio for RO + R203, SiO2 12. It is interesting to note that several relatively recent investigators have used similar ratios as the basis for their discussion of mica chemistry. A systematic attempt to explain the chemistry of the micas was made by Tschermak (1878). In brief, Tschermak regarded the micas as consisting of four fundamental molecules to which the following formulas were assigned: 1. R'6A16Si624 R' = KNa,Li or H 2. Mg6Si 624 27

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 3. H8Si10O24 4. F Si.0 24 10 8 Aluminum may be replaced by ferric iron and chromium; magnesium is equivalent to ferrous iron and manganese. The first of these formulas represents fairly well the composition of muscovite except for Tschermak's (1878) use of the double form of the molecule. Followers of Tschermak used the following formulas to represent the individual molecules: 1. R' 3A3Si3012 2. Mg6Si312 3. H4Si5012 4. F Si04 12 5~4 H2KA13Si3012 was given to represent the formula of common muscovite, which is similar to KA12(SiAl)4010(OH)2, the one used today. Other ratios between K and H, e.g., HK2A13Si3012 were also hypothesized. The chief disadvantage in the former notation is that the true relationship between the various atoms, e.g., substitution of aluminum for silicon or indivuality of the hydroxyl radical, is not recognized. The other Tschermak molecules represent hypothetical compounds. The second is a hypothetical polymer of chrysolite and the third is a hypothetical silicon hydroxide which may also take the form of the fourth, F12Si504. The last two were not analogous to any known substance, and as pointed out by Clarke (1889, p. 584), were chemically improbableo Tschermak believed that the micas were isomorphous mixtures of these molecules. According to Dana (1892, p. 612), Rammelsberg (1878, 1879) "Regards the micas as containing the three silicates R2SiO3, R4SiO4, R6SiO5 in various molecular relation, e.g., Muscovite is R4SiO4 + A14Si3012; the more acidic kinds are R14Si40O1 = R2SiOP + 3R4SiO4, which is further written mR14Si 415 + nRS + pR14Si12045, in which m:n:p = 5:1:5, 7:1:7, 9:1:9 in different cases. Similarly the other micas are resolved into the same three silicates, and the ratios in which they enter are calculated. That this method of calculation is applicable to any silicate, however complex, is obvious, but it is difficult to believe that the results reached really give the true constitution of the compounds." 28

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Tschermak's theory of the mica group remained generally in vogue and unchallenged until Clarke (1889) proposed an alternative. His hypothesis was considered by many to be superior. Dana (1892, p. 613) remarks, "The view of Clarke has the advantage that it assumes only one hypothetical molecule, which, moreover, is analogous to known compounds which play an important part in the Feldspar Group." All orthosilicates (SiO, radical) were considered substitution derivatives of the normal salt Al4(SiO4)3 in which the aluminum is replaced in varying amounts by univalent elements such as hydrogen, potassium, sodium, lithium, or univalent groups as MgF,A10, and AlF2. They are represented in the following formulas by R: Si0 - R 4 3 Al -SiO4 Al 1. R3A13(SiO4)3 Si = Al S iO4 R3 Si04 R Al-Sio 4 --- R 2. R6A12 (SiO04)3 SiO4 A. A/ - 4 3- Al-SiO4 R3 3. R9Al(SiO4)3 Si04 R3 The second of these formulas is not essential as it can be regarded merely as a mixture of equal molecules of the first and third types. Derivatives of the first type include R'R" A1 (SiO4) or R" Al (SiO4)6. R" represents magnesium, iron, manganese, etc. Likewise, erivatives of the second and third types may be obtained. All the orthosilicates may be explained by means of the first and third derivatives. Those micas with fluorine are represented by the univalent groups MgF or AlF2. Excess oxygen is represented by the univalent A10 group. In the case of some muscovites, phlogopites, and many lepidolites, the oxygen-silicon ratio fi low, which indicates a more acid condition. These micas are therefore considered to contain, in part, polysilieie acid, H4Si308, which, like orthosilicic acid, H4Si04, is tetrabasic.c This line 29

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN of reasoning, which is similar to the one used at that time to explain the plagioclase series, necessitates isomorphous miscilibity between orthosilicates and polysilicates (Si308 radical). Thus the general formulas for all micas may be obtained from the theoretical initial units A14(SiO4)3 and A4(Si308)3. Examples of the initial points of the orthosilicates and corresponding polysilicates are given below.: 1. mR3A13(Si04)3 nR Al3 (Si3O8 ) 2. mR6Al2(Si14) nR6A2 (Si38)3 3. mR9Al(SiO4)3 nRgAl(Si308)3 Clarke (1889) notes that if SiO4 and Si108 were represented by X, then all micas fall within the limits R3Al3X3 and RAlX3. Ordinary muscovite was represented by the first type, R3A13(Si04)3,| in which R = H2K, giving H2KA1(SiO4)1. Phengite, the high silica muscovite, was regarded as an isomorphous mixture of normal muscovite and the polysilicate H2KA13(Si308)3; other micas, such as lepidolite, are correspondingly much more complicated. Kunitz (1924) made the first noteworthy modern attempt to correlate the chemical composition of the micas with physical and optical constants. In the course of his work he made 32 chemical analyses of various micas, Three main isomorphous groups of the micas were advocated by Kunitz: 1. Alumina micas, or the muscovite series, in which aluminum is replaceable by as much as 10% of Fe3. 2. Magnesia-iron micas, or the biotite series, in which magnesium and ferrous iron are completely replaceable. 3. Lithia micas, or lithionitesi Here Kunitz assumes a special trivalent complex group of elements consisting of lithium and silicon in variable proportions (Generally 2Li, Si), to which he assigns the symbol Le. This group is completely replaceable by ferrous iron. Lithia micas, therefore, may be written as KH2Al2Le (Si04)3. 130

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Kunitz (1924) points out that members of the lithionite series that are rich in lithia contain little alumina, and vice versa, and recognizes that lepidolite can be derived from muscovite by the substitution of Li20 for A1203. He considers that there is a certain kind of isomorphous relationship between magnesium (and ferrous iron) and lithium. Partial isomorphous replacements also occur in the case of Si and Ti; Si and Al; Al, Cr, and Mn; K and Na; but not between K and H, or Al and Mgo His results show also that with increasing iron there is a corresponding, often linear, increase in the specific gravity and in the refractive indices. The refractive indices are to some extent modified by isomorphbus replacement of fluorine and hydroxyl. In 1925, Hallimond (1925, 1926, 1927) and Winchell (1925, 1927) presented almost simultaneously two new theories regarding the chemistry of the micas. In his discussion Hallimond (1925, 1926, 1927) emphasizes two main points, namely, that RO can replace R203, and that the invariant ratio of K20:SiO2 is in the proportion of 1:6. On the basis of these assumptions, calculations are made and ideal formulas for various species and end-members are given. The micas are divided into two main sections; the acidic, which is further divided into the potash and lithia micas, and the basic, which comprises the biotites and phlogopites. That RO can replace R20O, Kunitz (1924 po 389) denies on valency and theoretical grounds. Winchell (1925, 1927), however, believes that the RO:R203 ratio is variable and that K20 is constant, but cannot agree with Hallimond's (1925, 1926) interpretation of either of these facts. Winchell (1925, 1927) observes that the RO:R203 ratio varies because A120 replaces MgSiO3 (MgOSiO2)and vice versa. Winchell (1925, 1927) further regards potash as being constant and contends that not only does the K20:Si02 ratio vary, but that it varies in a regular manner from 1:6 in muscovite and phlogopite to 1:5 in protolithionite and siderophyllite and to 1:7 in phengite. One of the other points stressed by Winchell (1925, 1927) in his early studies of the micas is the belief that isomorphism depends on atomic volumes rather than on the valency of the elements which may proxy for one another (contrary to the opinion of Kunitz, 1924). In the micas, fluorine and hydroxyl may proxy for oxygen; potassium does not proxy for hydrogen, nor titanium for silicon. Winchell (1925, 1927) used many analyses to calculate end-members (many of which were revised in later works) of the biotite-phlogopite and muscovite-lepidolite series. In these contributions Winchell first used the terms heptaphyllite, (muscovite and lepidolite systems) and oatophyllite (biotite system) for the two main groups of micas between which there is no isomorphism. This is the result of the observation that micas fall into the two distinct classes, the fundamental units of which contain seven and eight positive atoms respectively. These terms, somewhat modified, have now been generally accepted. 31

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN In a series of ten papers with the general title, "Contributions on the Chemical Constitution of Micas," Jakob (1925A,B; 1926; 1927; 1928; 1929A BC,D; 1931; 1932) took an entirely different approach to the chemical and structural problems of the micas. The method is based on his theoretical and highly speculative theory of the constitution of the silicates (1920, 1921). A basic part of the theory is that certain atoms or groups of atoms are more closely united than others in a central atom, and may lie in an inner sphere of attraction while others are separately ionized and are possibly more remote. Jakob (1920, 1921) concludes that groups in the outer sphere have little or no influence on the capacity of two minerals to form mixed crystals. The constitution of the silicates he refers to certain prototypes [SiOR'41R [SiO]R'6 and LSiOj R'8 with or without added SiO2 groups. It now appears that this theory has no basis either structurally or chemically. With the application of X-ray diffraction technique to the study of natural micas between 1927 and 1932 (see section II of this part), new data was brought to light and with these new facts in mind Winchell (1932, 1935) revised some of his previous studies on the lepidolites and biotites. Lepidolites which had been classed as heptaphyllites are now, on the basis of Mauguin's (1927, 1928A, 1928B) X-ray results, regarded as octophyllite micas along with the biotites and phlogopites. New end-members in the biotite system were given as: annite 5(H4K2Fe6Al2Si6p24), siderophyllite 6(H4K2Fe5A14 Si 0.), eastonite 6(H K2Mg Al4Si 04 ), and phlogopite 5(H4K^Mg6Al Si6024). The end-members suggested by Winchell (1932, 1955) in these later investiations are in essential agreement with the conclusions of Berman (1937). In his final paper on the micas, Winchell (1942) made his fourth attempt (others1925, 1927, 1932) to correlate the chemical composition and optical properties of the lithia micas in which he took cognizance of the polymorphism established by Hendricks and Jefferson (1939). Newly calculated end-members include polylithionite (K2Li4Al2Si8o20oF), protolithionite (K2LiFe4Al3Si6020F4), and paucilithionite (K2Li3Al5Si602oF4). Winchell (1942) has also noted that in most analyses of lepidolite the Li2O content is deficient by almost one percent. This, he considers is due to the presence of some interlayered, but not isomorphously combined, muscovite. The work of Stevens (1938) is the only other important contribution to the chemistry of the micas before it was realized that the micas crystallize in many polymorphic forms. This work on the lepidolites resulted in the interpretation of seventeen new analyses on the basis of isomorphous mixtures between four arbitrarily chosen end-members. They are polylithionite (KLi2AlSiO4010F2), biotite K(Mn,FeMg) AlSi3010 (F,OH)2, lithium muscovite (K4Li6Al6A1 Si12 O0 (FOH )8), and muscovite (KA12AlSi01 (OHF)2). Volk (1939) studied the optical properties and chemistry of muscovite and suggested that the variations in composition and correlations with optical data may be interpreted with reference to the three end-members; 52 32

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN potassium muscovite, H4K2A16Si6024; phengite H6K2(Fe,Mg)2Al4Si6024; and ferric iron muscovite H4K2Fe32A14Si6024o The formula proposed for phengite is new and was used in order to get satisfactory correlation with the chemical analyses. The ratio of R20:SiO2 remains almost constant in all three members. The chemical constitution and classification of micas has also been discussed by Serdiuchenko (1948A). He proposed the idea that an almost continuous isomorphous series between muscovite and phlogopite is characterized by the replacement R32 —R23(Mg — A12) in the octahedral layers. The composition of biotite is considered intermediate between these two micas, ioe., isomorphism exists between heptaphyllites and octophyllites. This idea stems from the theory that in the biotite-phlogopite series, isomorphism takes place not only between magnesium and ferrous iron, or aluminum, ferric iron and chromium, but also between R32 and R23. This latter type "characterizes a limited isomorphism in the alternation... 6Mg(OH)2... and.. 4Al(OH)3... layers in the lattices of the minerals." (In translation.) The concept of alternating layers with octahedral coordination having different chemistry is similar, in a way, to that of Holzner (1936), who believed that biotite is intermediate in composition between phlogopite and muscovite. Furthermore, he suggested that the crystal structure of biotite is that of alternating layers of these two micas. Chemically he considered biotite as having the approximate composition of two phlogopites to one muscovite, which is KMg2Al0.667 (AlSi3010OH) 2 with the usual isomorphous replacements. Serdiuchenko (1948A), moreover, suggests broad isomorphous replacements in the tetrahedral layers after noting that the Si:Al ratio fluctuates widely from 2.4:1.6 to 3.6:0.4. Silica rich micas such as phengite and alurgite are explained on the basis of the relatively large silicon content in these sheets. Selected biotites are cited as examples of micas intermediate in composition between biotite and muscovite. Serdiuchenko (1948A) also states (in translation), "In connection with the presence of polymorphic modification in the group of micas, the points of the heterostructural micas may be ranked on the line of our diagram." This is interpreted as implying that the various polymorphic forms of biotite, for example, are direct consequences of their chemical position between muscovite and phlogopite. From this discussion of the chemical history of the micas it is clear that the end-member concept is firmly engraved on the minds of many mineralogists. The requirements necessary for end-members are indicated by Winchell (1925, 1942) and Stevens (1938, 1946), and a method for calculating the ratios of end-members in the micas is well explained and illustrated by Stevens (1946). Heinrich (1946, p. 846), however, has pointed out that 33

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN "The theoretical'end-molecules' used by Kunitz (1924) and Winchell (1935) have little significance in the structure of the biotite molecule, nor can variation in chemical composition of the series be expressed adequately in terms of them," Rabbitt (1948, p. 284), in his work on the anthrophyllite seriesi concurs and notes that the molecular end-member concept is founded on false premises "No such end-members in the amphiboles are known in nature. No such mixing on a molecular basis occurs. However convenient it may be to plot the different properties in that way it must be admitted that it engenders and perpetuates a false notion of the variations in a mineral series." B. Chemical-Structural Systematization of the Natural Micas A workable systematization of the micas has long been one of the aims of the project. A classification based on structural and chemical data, along with an account of the very voluminous synonomy of the micas, has been completed. In the following pages this classification is presented. The numerous attempts at mica classification in the past have been based almost entirely on their chemistry and fall generally into two categories: (1) those employing complex quasistructural formulas (e.g., Clarke, 1889; Jacob, 1925); and (2) those involving calculation of theoretical endmember molecules (e.g., Tschermak, 1878; Kunitz, 1924; Stevens, 1946). Berman (1937) approached the problem in more realistic fashion by correlating mica formulas with unit cell contents and expressed the chemical variation by means of atomic ratios of different elements occupying equivalent structural positions. However, the work of Hendricks and Jefferson (1939), Levinson (1953), and others has demonstrated the complex polymorphism in the mica group and the necessity for both chemical and X-ray studies of micas as a basis for a sound classification. Because of the complex chemistry of the group, the numerous textural and other minor varieties, and the repeated misapplication of names, the nomenclature and synonomy of the micas have been ponderous and involved. The purpose of this work is threefold: (1) to present a reasonable and usable subdivision of the micas into naturally occurring species and major varieties on the basis of both chemistry and structure; (2) to review the synonomy of the group and thus assign minor varietal names or duplications to the species or major variety to which they correspond; and (3) to list those micas that are of indeterminate status owing to incomplete studies. ----------- ^ -----------

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN In the annotated classification that follows we have used for chemical varieties the prefixes of Schaller (1930). Wherever possible for each entry we listo (1) a simplified formula based on 24 negative ions in the unit cell; (2) the major isomorphous substitutions; (3) the structure; (4) notes; and (5) synonymy. For the structural varieties, the method of polymorphic notation suggested by Ramsdell (1947) is adopted. The symbols and their meanings used in this work are as follows: 1M 1-layer monoclinic structure 2M muscovite 2-layer monoclinic muscovite structure 2M 2-layer monoclinic octophyllite mica structure 3M 3-layer monoclinic structure* 3H 3-layer hexagonal structure 6M 6-layer monoclinic structure 6T 6-layer triclinic structure 24T 24-layer triclinic structure 18T 18-layer triclinic structure (Amelinckx, 1952B) The chemical-structural classification of the micas is as follows. Species 1. Muscovite, K2A14(Si6Al2)020(OH)4. Minor NaRb,Cs,Ba, and Ca for K; minor MgFe2,Mn2 and Li; Minor Fe3,Ti, and Cr for Al; minor F for OH; maximum Li20 = 3.30, occupying vacant octahedral positions. Structure: 2M muscovite. Synonyms: adamsite, ammochrysos, amphilogite, biaxial mica, cat gold. cat silver, common mica, damourite (in part), didrimite, did-ymite, ferroc-ferri-muscovite, frauenglas, helvetan, heptaphyllite, isinglass, kaliglimmer, marienglass, morrepit nacrite, oblique mica, potash mica, schernikite, zwefxi'ger glimmer. Hypothetical end-members: ferri-muscovite, kryptotile, leverrierite, lithium muscovite. Varieties. a. Barian muscovite. Ba with reported maximum of 5991% BaO (Doelter, 1914) for Ko Structure: probably 2M muscovite, material labeled oellacherite from Tyrol has the 2M muscovite structure. *There is reason for believing that the 3-layer monoclinic structure reported by Axelrod and Grimaldi (1949) is not truly monoclinic but just a distorted 3-layer hexagonal structure., This point is explained in further detail on page 47. 35

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN The barium-muscovite from Franklin Furnace, New Jersey, described by Bauer and Berman (1933), was examined by us by means of the X-ray powder method. The resultant pattern is not that of a muscovite, nor even that of a mica. The material is too finegrained for detailed optical studies (See p. 68). Synonyms: barium muscovite, oellacherite, sandbergerite. b. Manganian-muscovite. Reported maximum MnO = 2532% (Ellsworth, 1932); usually some Li20 present. Structure: 2M muscovite. The very fine-grained,- deep purple mangan-muscovite of Eskola (1914) with 2.35 MnO has been checked by means of X-ray powder photographs, and the pattern does not correspond to any known mica structure. The photograph shows some quartz lines; see page 69 Synonyms: mangan-muscovite, manganese muscovite. c. Ferrian muscovite. Fe+3 with maximum of 5.70% Fe203 (Tschermak, 1878) reported, for Al. Structure: probably 2M muscovite, for the "alurgite" from Cajon Pass (Webb, 1939) with 5.32% F-e20 has been shown to have this structure. d. Ferroan muscovite. Fe+2, with maximum of 6.55% FeO (Wulfing, 1886) reported. Structure: probably 2M muscovite. e. Chromian muscovite. Reported maximum of 4.81% Cr203 (Whitmore, et al., 1946), for Al. Structure: 2M muscovite. This follows the usage of Whitmore et al. (1946). Synonyms: chromglimmer, chrome glimmer, fuchsite, gaebhardite, verdite. f. Lithian muscovite. K2(Al,Li) ca. 5.0 (Si6 7,A1 1) 020(OH + F)4. Li2O, at least 3530% occupying vacant octahedral positions. Usually small amounts of F for OH. Structure: modified 2M muscovite (Levinson, 1953). g. Phengite. K2(Mg,FeAl)4(Si7,Al1)020(OH)4. High-silica muscovite with considerable MgO (7.96% Pagliani, 1937) and in some cases FeO; some Fe3 for Al. Structure: 2M muscovite. ------------- ^~~3

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN In addition to the 2M muscovite and 3H types of phengites, Dr. H. S. Yoder has informed us that some of the sericites from Amelia, Va. described by Glass (1935) have the 1M structure. The nomenclature of the silica-rich end of the muscovite series as proposed by Schaller (1950) is regarded by us as unsatisfactory because: (1) there is some evidence that the original leucophyllite (Starkl, 1883) is a mixture, and in any event a new analysis and an X-ray study of the type material are needed to check its validity; and (2) the term, alurgite, which Schaller (1950) suggests as a substitute for leucophyllite in the event that the latter should prove untenable, also has been erroneously employed to indicate a normal, i.e., low-silica, ferrian muscovite (Webb, 1939; Odman, 1950). X-ray studies, however, indicate that not all true alurgite is two-layer monoclinic in structure; some is three-layer hexagonal. Thus the term alurgite has at various times been used for: (i) a red high-silica muscovite with minor Fe3 and Mn (Penfield, 1893); (ii) a red ferrian low-silica muscovite (Webb, 1939 and Odman, 1950); (iii) a manganian high-silica muscovite (Winchell and Winchell, 1951); and (iv) a three-layer hexagonal polymorph of (i), discovered by Hendricks and Jefferson (1939) and verified by us on material from St. Marcel, Italyo Because of this confusion, the use of the term alurgite for the high-silica end-member of the muscovite series is also undesirable. Less confusion accompanies the term phengite, which has been generally employed to mean high-silica muscovite. Hypothetical end-members: ferrophengite, picrophengite. h. Hexagonal phengite (3H phengite), Differs from 2M phengite (g) in having the three-layer hexagonal structure (3H) and 2V = 0~ - low. No well authenticated analysis of all-uniaxial material is available. i. Chromian phengite. Cr, with maximum of 0.78% Cr203 (Whitmore, etoal., 1946), for Al. Structures 2M muscovite. This follows the usage of Whitmore et al. (1946). 37

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Synonyms: chromochre, mariposite. j. Hexagonal chromian phengite (3H chromian phengite). Presumably chemically similar to ii, but with the three-layer hexagonal structure and 2V = 0~ - low. No analysis of alluniaxial material available. k. Sericite. Fine-grained muscovite. Structure: 2M muscovite. The term-sericite has been used for: (i) -fine-grained muscovite, either primary or secondary; (ii) fine-grained phengite; and (iii) hydromuscovite. Thus it cannot be defined exactly on a chemical basis. Iti;' remains, however, highly useful as a general, nonspecific term for fine-grained muscovite whose exact chemical nature is unknown. Synonyms: (including pinitic pseudomorphs). achlusite, agalmatolite, aspasiolite, avalite, bildstein, bonsdorffite, catalinite, cataspilite, cordierite-pinite, cymatolite, damourite (in part), dysyntribite, epileucite, epi-sericite, fahlunite, giesickite, gigantolite, gilbertite, glimmer, gongylite, helvetan hygrophyllite, iberite,:ivgitite, killinite, lardite, lepidomorphite, liebenerite, lythrodes, margarodite, micarel, micarelle, oncophyllite, oncosine, onkophyllit, onkosin, oosite, pagodite, parpphite, pinite, pinitoid, polyargite, pyknophyllite, pyrargilliter pyrrholite, rosellan, rosellite, rosite, shilkinite, sterlingite, talcite, terenite, triclasite, wilsonite. Species 2. Paragonite (Na,K)2A14(Si6,A12)020(OH)4. Structure: 2M; probably 2M muscovite. Schaller and Stevens (1941) have pointed out that the series muscovite-paragonite is not completely represented in nature. If intermediate types are discovered, it might be better to regard paragonite as sodian muscovite and reduce it to varietal status. Synonyms: hallerite, natronglimmer, pregrattit, soda mica. Species 3. Roscoelite. K2(V,Al)4(Si6A12)20p(OH)4. Maximum V205 = (ca.) 20%. Structure: 1M. 58

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Although roscoelite has previously been regarded as a vanidiferous muscovite, it deserves full species rank because it is structurally distinct (see p 212 ), - Synonyms: colomite, vanadinglimmer, vanadium muscovite. Species 4, Lepidolite. K2(LiAl) 6(Si,onAl )O l(FOH)t Rb and Cs replace K, in some types in considerable amountsRb Cs20 = 1.90%; Lunblad, 1942). Small amounts of Mn,Mg,Fe2,Fe are normally present. The OH:F ratio varies considerably, and OH may become negligible. Synonyms: irvingite, lilalith, lilalite, Li-phengite, lithia mica, lithionglimmer, lithionit, lithionite, lithionitesilicat, macrolepidolite, microlepidolite, poly-irvingite, scale stone, s ideris cher-fels -glimmer. Hypothetical end-members: paucilithionite,' polylithionite (in part), protolithionite. Varieties a. a.' Six-layer: m6onoclinic' l6pidoiite (6M lepidolite). Li 0 = (ca.) 4.0 - 5.1%. b. One-layer monoclinic lepidolite (1M lepidolite). Li20 = 5.1- 7.26%. c. Three-layer hexagonal lepidolite (3H lepidolite). Composition approaches 4b. Due to twinning (?) (Levinson, 19553). 2V = O -small'. d. Three-layer monoclinic (3M lepidolite). Li20 = 4.1%. One example from Skuleboda, Sweden (see p. 147 ) * n e. Manganian-lepidolite, Maximum reported MnO = 7.55% (Shibata, 1952B). Structure: probably variable, depending on Li content. f. Magnesian-lepidolite (cited in Berman, 1937). We are unable to determine if any natural material of this composition has been discovered. g. Polylithionite (in part). K2Li4A12Si8020(F,OH)4. A silicon- and lithium-rich, thus aluminum-poor, lepidolite. Structure: 1M. Species 5. Taeniolite. K2Mg4Li2Si80O20F4 Structure: 1Mo 39

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - Species 6. Zinnwaldite. K2(Fe21_2Li2_A12) (Si6-7,A12_-1 020 (F53-20 Hl-2). The (AlFe,Li) group usually contains 6 atoms but may be considerably deficient. Zinnwaldites are chemically much more closely related to the biotites than to the lepidolites. In fact, there is no sharp natural compositional boundary between zinnwaldite and lithian biotite. If a demarcation is to made, we believe the line can best be drawn on the amount of Li and suggest that lithium-iron micas with Li> 1 atoms per unit cell formula be regarded as ziannaldites,those with Li < 1 be placed with the lithian biotites. Structurally, however, the problem is more complex. No zinnwaldites have been found with the 6M lepidolite structure. so common in the lepidolites. On the other hand, the discovery of a zinnwaldite with a structure very similar to that of lithian muscovite (p. 201 ) does not permit one to overlook the possibility of at least some zinnwaldites resembling lepidolite structurally. Synonyms: cryophyllite (in part), lithioneisenglimmer, polylithionite (in part), protolithionite, rabenglimmer. Varieties. a. One.layer monoclinic zinnwaldite (1M zinnwaldite). b. Two-layer monoclinic zinnwaldite (2M zinnwaldite). c. Three-layer hexagonal zinnwaldite (3H zinnwaldite). d. Ferrian zinnwaldite maximum report Fe203 = 10.06% (Shibata, 19523). Species 7. Phlogopite. K2(MgFe2)6(Si6A12)0Oo(OH)4. Na can substitute for K up to nearly K:Na =1:1 Harada, 1936); minor RbCs, Ba, and very minor Ca also may proxy for K. Fe is almost always present, but Mg predominates greatly over Fe2. Small amounts of Mn,Fe3 and Ti may be present. The Si:tetrahedral Al ratio may be larger than 6:2. There is no well-defined, natural, compositional boundary between ferroan phlogopite and magnesian biotite. Because Fe2 is a strong chromophae, micas of this type even with only a small per cent of Fe are dark colored and are thus commonly classed as biotites. If a division is required, we suggest that where the ratio of Mg:Fe2> 4:2, the mineral should be classed as phlogopite. _40

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Synonyms: aspidolite, barium-phlogopite, barytbiotite, hydrophlogopite, magnesia mica, natronophlogopite, octophyllite, pholidolite, rhombic mica. Hypothetical end-members: fluor-phlogopite, hydroxylphlogopite. Varieties. a. One-layer monoclinic phlogopite (1M phlogopite). The most common type. b. Two-layer monoclinic phlogopite (2M phlogopite). c. Three-layer hexagonal phlogopite (3H phlogopite). d. Manganophyllite. K2(Mg5_4,Mn21 Fe2 0oFe3_ 0-1) (Si6Al2)0O0( 0H)4. Structure- Generally 1M; some approach a three — layer hexagonal structure. 2M reported by Hendricks and Jefferson (1939). Although some investigators (e.go, Hey, 1950) class manganophyllites as varieties of biotites, most manganophyllites have little or no Fe2 and only small amounts of Fe3. An exception is a Langban, Sweden, mica analyzed by Jakob (No. 8, po 157, 1925A)which contains 16.941 Fe203o Apparently Mn is present commonly as Mn3, rarely as Mn2o Synonyms: Manganese mica, manganophyllo e. Titanian phlogopiteo K2Mg5Ti(SiAl12)020(OHF)2 (Prider, 1940) and Fe2 and Fe3 are minor; (OHF) is deficient; 0.15 atoms of Ti are allottedto tetrahedral positions and 0.84 atoms to octahedral positions on basis of original analysiso Structure: 1M1 Species 8. Biotite. K2(Fe2,Mg)64(Fe3,AlTi)O2(Si65,Al2-)0222(OHF)4-2 Some Na,CaBa,Rb, and Cs for K; Mn for Fe2, F for OH; Mg may be almost absent. Total F + OH may be very low (Walker and Parsons, 1926). Synonyms: annite, anomite, caesium-biotite, chromglimmer (in part), euchlorite, eukamptite, ferromuscovite, haughtonite, heterophyllite, hexagonal mica, iron mica, lepidomelane, meroxene, natronbiotite, octophyllite, odenite, odinite, odite, oderite, pterolite,.hombenglimmer, rubellan, siderophyllite, titanglimmer, titanmica, uniaxial mica, waddolte. 41

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Hypothetical end-members: cryophyllite (in part), eastonite, fluor-annite, fluor-biotite, fluor-lepidomelane, fluor-meroxene, fluor-siderophyllite, hydroxyl-annite, hydroxyl-biotite, hydroxyllepidomelane, hydroxyl-meroxene, hydroxyl-siderophyllite, manganophyllite (in part), Varieties. a. One-layer monoclinic biotite (1M biotite). The most common type. b. Two-layer monoclinic biotite (2M biotite). c. Three-layer hexagonal biotite (3H biotite). d. Six-layer triclinic biotite (6T biotite). Our studies of the supposed type 6T biotite described by Hendricks and Jefferson (1939) from Sterling, N. Y. (U.S.N.M.C3675) indicates the 1M structure. Dr. H. S. Yoder has studied other material from the same specimen and has found the 2M structure. e. Twenty-four-layer triclinic biotite (24T biotite). f. Eighteen-layer triclinic biotite (18T biotite), Amelinckx (1952B). g. Calcian biotite. A biotite from Kaiserstuhl, Germany, has 14.335 CaO; Zambonini (1919). Structure unknown. The validity of calcian biotite as a major variety is doubtful. The existence of Ca in the biotite structure has been challenged by Jacob (1929A). Several specimens labelled calciobiotite, from Italian localities, have been found to have the 1M structure. h. Ferroan biotite. Mg is very minor or absent, Fe is present mainly as Fe2. Synonyms: siderophyllite, lepidomelane (in part). i. Manganian-biotite. Mn as much as 1 atom per unit cell formula. Fe present as Fe2 or Fe5. j. Ferrian biotite. K2(Fe2 Mg)3 4Fe32_ (Si,AlTi)8020_21 (OR,F)43. Synonyms: ferribiotite, lepidomelane (in part). ( 4 Lepidomelane is commonly employed for iron-rich biotite, but the term has been used to embrace biotites rich in Fe3, those rich in Fe2 and those with relatively large amounts of both Fe2 and Fe3 (Heinrich 1946), -----------------------— 4 2 --------------

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN k. Lithian biotite. K2 (Fe2,Mg)5 (Li,Al,Fe35)(SiAl)802 (OHjF)4. 1. Titanian biotite. K2(Fe2,Mg3)5Ti(Si6,A12)8022 (OHF)2. Synonyms: ferrititanbiotite, ferriwotanite, titanbiotite, titanobiotite, wodanite, wotanite. Freudenberg (1919) reports a biotite with 12.5% TiO2. Both titanian biotite and titanium pholgopites are relatively poor in Fe3 and also are very low in OH and F. Hydrous Micas. Because most investigators class these minerals with the clay group, their crystal chemistry is not considered here. However, it is interesting to note that a new interpretation (Brown and Norrish, 1952) of the chemistry of one of the species in this group, hydromuscovite, postulates the replacement of K by oxonium (hydronium) ions H30+. Species and varietal names included in this group are: bastonite, brammallite, bravaisite, buldymite, damourite (in part), goeschwitzite, grundite, gumbelite, hydrobiotite, hydro-mica, illite, metasericite, Mg-illidromica, rastolyte, sarospatakite, sarospatite, sericite (in part), sodium-illite, voigtite. Micas of Indeterminate Status. 1. Euphyllite - near (Na,K)A3SiOl o(OH)2; may be a mica intermediate between muscovite and paragonite; or brittle mica or perhaps a mixture. 2. Mahadevite -near (K,Na)o.97(Al,Fe,Mg)2.66(Si,A1)4(O,OH)2; supposedly between muscovite and phlogopite in composition, A specimen of mahadevite has been received from Dr. H. S. Yoder. X-ray studies of this material indicate the 1M phlogopite structure. 3. Manandonite - a borosilicate of Li and Al closest to lepidolite in composition; Li4A114B4Si60O29(OH)24(?); possibly- not a- mica. 4. Leucophyllite - a high silica phengite; perhaps a mixture. 5. Anthrophyllite - a "mica (?)" (Hey 1950, p. 283). L —-------------------— 45 --------------

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN PART II. MUSCOVITE I. MINERALOGY OF NORMAL MUSCOVITE A. Chemistry The ideal formula for muscovite is K2A14(Si6Al2)020(OH)4. However, all micas exhibit a large number of isomorphous replacements. The following isomorphous substitutions are of major proportions: for K —Na,Rb,Cs,Ba, and (Ca) for octahedral Al —Mg,Fe2,Fe3,Mn,Li,Cr,V and Ti for OH —F for (Si6A12) —the proportion may approach (Si7All) as in phengite For normal pegmatitic muscovite the percentages of isomorphously substituted elements may reach relatively large proportions. Many such micas with over 4 percent total iron, for example, have been reported or are now known on the basis on new spectrochemical data. It appears that rose muscovites, which are the last micas to crystallize in the hydrothermal sequence of complex pegmatites, approach most closely to the ideal muscovite composition (Heinrich and Levinson, 1953)o They contain very little Fe2,Fe3,Mg, or Mn, etc. Schaller (1950) has pointed out that the high-silica sericites (phengites) contain appreciable amounts of a divalent element, usually Mg* Pagliani (1937) reports a phengite with 7.96 percent MgO. It has been reported by several authors (Gruner, 1948 and Sudo, 1949) that the fine-grained micaceous minerals which go under such names as illite, -hydronica, sericite (in part) and hydromuscovite commonly show a deficiency of K and an excess of H20. Sudo (1949) showed that the (OH) content is roughly inversely proportional to the combined content of K,Na, and Ca. Gruner (1948) postulated that the hydronium ion (H30+) may occupy the vacant K portions, and Brown and Norrish (1952) supported this idea further by means of calculations. Several occurrences in which large books of normal pegmatitic muscovite are replaced by fine-grained sericite aggregates may be explained by a hydronium sericitic mica replacing a potassium muscovite (Liashchenko, 1940). 44

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN For this study 168 muscovites were analyzed spectrochemically. The results are shown in Table I. TABLE I NEW SPECTROCHEMICAL DATA ON 168 MUSCOVITES Present in % Range Average of samples Fe20 100 0.63 - 5.1 2.904 Mg 100 0.008 - 1.9 0.931 MnO 100 0.009 - 0.83 0.106 Ti02 99 0.003 - 0.52 0.128 CaO 100 0.0005 - 0.43 0.0069 SrO 21 0.001- 0.005 0.0005 BaO 96 0.0002 - 1.1 0.114 Na20 100 0.33 - 1.7 1.02 Li20 26 0.01 - 1.6 0.089 Rb2O0 O Cs2Q 0 Ga203 100 0.003 - 0.062 0.0219 Sc203 31 0.001 - 0.011 0.0011 Co203 37 0.0002 - 0.0008 0.0002 Cr203 32 0.0001 -.4 0.0001 SnO2 76 0.0007 - 0.10 0.0067 V205 65 o.0001 - o.o36 0.0004 F 44 0.10 - 1.48 0.2361 B. Muscovite Structure 1. General and Polymorphism. Jackson and West (1930, 1933) made the first detailed study of the structure of mica (muscovite) and, in addition, confirmed in greater detail the structure of mica and other layered silicates proposed by. Pauling (1930) on the basis of his coordination theory. The results obtained by Jackson and West (1930, 1933) on the unit cell dimensions of muscovite are as follows: ao = 5.18 A, bo = 9.02 A, co = 20.04 A.

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - They selected 95~30' as the P angle in contrast with about 100~ chosen by Mauguin (1927, 1928A, 1928B). The monoclinic space group determined by Mauguin as C2/c was verified. The most detailed study of the mica group as a whole was conducted by Hendricks and Jefferson (1939) who found numerous polymorphs. In all, seven different polymorphic modifications embracing the hexagonal, monoclinic and triclinic systems were found among 100 investigated specimens of named micas. Each species, with the exception of muscovite, was found to occur in more than one polymorph. Biotite, for example, appears in five modifications, lepidolite in four, etc. Muscovite, however, Hendricks and Jefferson (1939) found only as the 2-layer monoclinic muscovite type (2M muscovite). They also found one "lepidolite" and one alurgite with this structure. The distinguishing feature of the muscovite structure is its distortion from that of the ideal mica. This is revealed by the presence of the reflections, (06G)with e odd, which should normally be absent in the ideal arrangement on the basis of the structure-factor calculations of Jackson and West (1930, 1933); see Fig. 5. This distortion results from an incomplete filling of the octahedral positions and is considered by Hendricks and Jefferson (1939) to be the factor permitting only the 2-layered structure for muscovite to form. Herein also lies the reason why Winchell's (1925, etc.) grouping of the micas into heptaphyllite and octophyllite divisions appears correct, for these terms simply mean that the unit cell contains seven and eight atoms respectively, for twelve 0,OH, and F. Muscovite, KA12(AlSi3)010(OH,F)2, is the type heptaphyllite, and phlogopite, KMg3(AlSi3)010(0HF)2, is the type octophyllite. This relation may also be expressed by saying that the heptaphyllites have only 2/3 of the octahedral positions filled, whereas the octophyllites have all such positions filled. The presence of(063ereflect ions with e odd must imply departure from the ideal muscovite structure given by Jackson and West (1930, 1933). Hendricks and Jefferson (1939) reported these reflections absent in the two-layered biotite-like micas (2M). It follows therefore that muscovite (two-layered) has a structure different from that of the two-layered biotites. Diffuse scattering is reported in all the micas except muscovite. The scattering is observed in Weissenberg photographs along those reciprocal lattice lines in which h and k are constant, but in which k is not divisable by three (hakbl, kbi3n). Laue photographs taken normal to the cleavage demonstrate scattering in the form of asterism or radial streaks. Mauguin (1927, 1928A,1928B) first noticed this effect and its significant absence in muscovit and theorized that it was the result of some type of randomness in the positions of the heavy ions. Hendricks and Jefferson (1939) explained it in terms of constant h and k indices with an apparently continuous variation of the ~ index, _____________________46 ___________

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN This results from a variable periodicity in the stacking of mica layers in such a manner that those planes in which the k index is a multiple of three are undisturbed. They further state that if one half of a mica layer is translated by nb/3 with respect to the other half, it leaves the layer unchanged with respect to itself but results in a change of the successive layers This is the: factor which permits polymorphism in the micas (but in the muscovite structure distortion prohibits translation)O Therefore, continuous scattering arises from destruction of the lattice periodicity perpendicular to the cleavage, resulting from a translation of some layers along the b-axis, parallel with the cleavage, by nb/3. Scattering has also been reported in other layered silicate minerals such as stipnomelane, vermiculite, cronstedite, and the chlorites. Essentially the same explanation is given for the phenomenon in these minerals. In our work we have found at least one example of an apparently normal pegmatitic muscovite with small amounts of diffuse scattering (Fig. 2; compare with Fig. 5). The scattering, however, is observed only along the 02e reciprocal lattice line (only O-level a-axis photographs have been taken of most muscovites). Scattering of the type described in the preceding paragraphs would require scattering along both the 02 and 042 reciprocal lattice lines. The exact significance of this type of scattering has not yet been determined. Noteworthy perhaps is the fact the specimen (Noo 6) from which the photographs of Fig. 2 were obtained has a SiO2 content of 48 percent which is about 3 percent above normal. Phengites and other high-silica muscovites, as will be shown later in this report are known to crystallize as more than polymorph: the two-layer normal muscovite and the 3-layer hexagonal. It is suggested that specimen No. 6 may be transitional between normal muscovite and the high-silica muscovites that crystallize in several polymorphs, and the diffuse scattering may be an embryonic attempt at a stacking arrangement other than the 2-layer type. In a recent paper, Axelrod and Grimaldi (1949) have reported a new polymorph of muscovite that contains three layers in the monoclinic unit cell and has a very small, variable 2V of from 3~ to 12~. A portion of this analyzed material has been obtained, and a re-study of it substantiates the existence of this new polymorph. However, there is a striking similarity between the "3-layered monoclinic muscovite" (space group C2 determined by Axelrod and Grimaldi, 1949) Weissenberg photographs and those of the 3-layered hexagonal mica polymorph (space group C3112 or C3212 determined by Hendricks and Jefferson 1939) recorded for biotite, alurgite, phlogopite, zinnwaldite and lepidolite. There is apparently room for a difference of opinion in the interpretation of the X-ray photographs of this muscovite. Axelrod and Grimaldi (1949) chose the former space group because of the presence of diffuse scattering in the (hakal) zone lines of one of the pseudo a- and pseudo b-axes, for they state (p. 569): -------------------— 47 -----------

oH'..........~ ~~~~~~~~~~~~~~~~~~~~IJ I~)~ ~~~~~~~~~~~~~~~~oc~ ~~~~~~~~~~~~~~~o ~~ kj1 ~ ii~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~........ ~.0 0' 0z 0 0 0~~~~~~~~~~~~ 00 c~~~~~~~~~~~~~~~~~~+C i~. Ii. o~~~~~~ H *'x,... —............................. -.............................. t.t t t tXHO co............................O ~ 0-............. ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ D~

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN "The X-ray symmetry, based on the symmetry of the individual Weissenberg patterns, was C2h or D3d. If the differencesin diffuseness between Weissenberg patterns are neglected, a threefold axis is demonstrated and the Laue symmetry is D3do If, as we prefer, the optic axial angle of 12~ is not ascribed to strain and the differences in diffuseness are not neglected, the Laue symmetry must be taken as C2h with the structure very close to trigonal." Many Weissenberg photographs of this polymorph have been taken and.studied by the writers in an attempt to assign definitely a space group to this muscovite. No distinct differences in diffuse scattering along the reciprocal lattice lines described by Axelrod and Grimaldi (1949) were found. Physical distortion or corrugation of the crystals X-rayed may produce effects that could outweigh the diffuse scattering inherent in the crystal structure. Although Axelrod and Grimaldi (1949) consider strain unlikely, we are inclined to favor this as the cause of the 2V of 12~. However, it is our opinion that the significant fact concerning this mica is that it contains 5 layers in the unit cell and not the fact that it is monoclinic because of its 2V of 12~, or because of diffuse scattering. We are convinced that if measurements are made with sufficient accuracy, many micas will be shown to be triclinic by a few minutes or seconds. On the basis of personal communication with several investigators we have every reason to believe that the P angle of the 3-layer lepidolite described by Hendricks and Jefferson (1939) varies slightly from 90~ and thus must be considered as having a symmetry no higher than monoclinic. (We have been informed of v angles of 90005' by Dr. J. Smith of the Geophysical Laboratory and 90~30' by Mr. S. A. Forman as measured on photographs from Hendricks and Jeffersonts original material). Hendricks and Jefferson (1939) note that some of the micas they describe as hexagonal show a small 2V and that all give some diffuse scattering. They state, however, (p. 762), "There is no doubt but that the hexagonal description is accurate as a limiting case." We agree with this statement and believe that the monoclinic nature of these 3-layer micas is probably due to physical defects and is not inherent in the mica structure. It may be practical, in the future, if precision of measurement continues to increase, to class and discuss micas merely as 1-layer, 2-layer, 3-layer, etc., without regard to crystal system, inasmuch as all multi-layer forms might be derived from the single-layer form by the application of various combinations of symmetry operations. Postel and Adelhelm (1944) have described a white muscovite of late hydrothermal origin from the Wissahickon complex in which 2V varies from 22~ to 50~. The explanation advanced is that random shift in the structural planes of mica may have some bearing on the low and variable 2V. In their paper, they 49

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN include a Laue photograph showing asterism which, if not due to some physical distortion, may indicate the existence of another muscovite whose structure is aberrant'. A study of the crystal structure in muscovite, as well as in many other micas, was one of the principal objects of this study. The Weissenberg method was used primarily and usually one photograph about the a-axis was sufficient to identify the polymorph. The structures of about 150 muscovites, chiefly of pegmatitic origin, have been identified and include "normal" muscovites, rose muscovites, sericites and several uncommon varieties such as pinite, margarodite, adamsite and alurgite. Almost all have the 2-layer monoclinic muscovite structure described by Jackson and West (1930, 1933) and Hendricks and Jefferson (1939), except: 1. The 3-layer polymorph of Axelrod and Grimaldi (1949) 2. Some alurgites and other high-silica muscovites and phengites 3. Lithian muscovite 2. Morphology. After X-ray studies revealed the prevalence of polymorphism in the micas it was evident that the long standing differences of opinion as to the correct axial ratios and P angle were a natural result of studies by investigators unaware of the fact that there are approximately as many correct interpretations as there are polymorphs. For almost a century, mineralogists had tried to obtain one set of constants applicable to all the micas, or in other words, to force unimorphism onto a group now known to be characterized by polymorphism. Muscovite is the only mica for which a set of constants may be obtained, as it alone, for all practical purposes, crystallizes in but one polymorphic form. Hendricks and Jefferson (1939) showed how the 3 angle, for instance, varies with the number of layers in the unit cell. If structural identity is contained in one layer, than the P angle will be 1000 (approximately); if two layers are required, the P angle will be 95 and will be 90~ if there are three or more layers in the unit cell. With these problems in mind, Peacock and Ferguson (1943) reviewed the whole field of morphology of the micas and muscovite in particular. In the following discussion constant reference is made to their work. Goniometric measurements of lepidolite and other micas are few compared with those for muscovite, and for that reason they are practically omitted in the following discussion. However, a complete understanding of the morphology of mica will be forthcoming, as the problems and methods are essentially the same as those for muscovite. Before attempting to explain the various crystallographic settings used in the past it would be best to present again the approximate axial ratios for the more common mica polymorphs based on unit cell dimensions determined by X-ray methods. 50

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN One-layer polymorph ao = 5.35 bo = 9.2A co = 10.2A P = 100~ axial ratio a:b:c = 0.58:1:1.11 Two-layer muscovite ao = 5.35 bo = 9.0A co = 20.0 A = 95030' axial ratio a:b:c = 0.58:1:2.22 Three-layer ao = 5.3A bo = 9.0A co = 30.0A P = 90~ axial ratio a:b:c = 0.58:1:3.26 The first crystallographic settings in general use were those of Miller. Peacock and Ferguson (1943) note: "In Brooke and Miller (1852, p. 389) biaxial mica, as distinct from uniaxial or nearly uniaxial biotite is described as monoclinic, with calculated angles which give elements that compare with the ratios of the singlelayered monoclinic lattice cell: a:b:c = 0.5774:1:1.1021 = 99058' Miller a:b:c = 0.58:1:loll = 100 Structure Thus Miller chose elements which are appropriate to single-layer micas, and may perhaps correspond to the structural type of the Vesuvian crystal which yielded the measurements, by Phillips, that were adopted by Miller. However, Miller's setting is unsuitable for the two-layer muscovite structure." Des Cloizeaux (1862, p. 485) developed the most widely used settings for the micas. They were adopted by Tschermak, Dana, Groth, Goldschmidt and others. Dana's (1892) values are: a:b:c = 0.57735:1:3.3128 P = 89~ 54 1/6' These values are in very close agreement with the values of the 3-layer polymorph which again are not suitable for muscovite and most micas. However, Tschermak (1877, p. 43) gave a special setting for muscovite from which the following ratios were computed by Dana (1892, p. 621): a:b:c = 0.5775:1:2.2175 P = 84~ 55' These values are in general agreement with the established constants based on X-ray studies. Even though this setting had certain advantages in simplicity of indices, Tschermak (1877) finally discarded it and returned to the incorrect settings of Des Cloizeaux (1862) in order to be consistent with other measurements. In an attempt at justification, Dana (1892, p. 620) remarks: 51

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN "The position here taken is that finally adopted by Tschermak (priv. Contro, May 1891) as the most satisfactorily showing the relation to other micas, the chlorites, etc." Kokscharov (1875), after studying many measurements, came to the conclusion that all the micas were monoclinic witha p angle of exactly 90~. He remarks (p. 235): "Alle Glimmer uberhaupt, ohne Ausnahme, wie die Optiker es behaupteten, zu dem rhombischen System mit einem monoklinischen Typus, oder riphtiger —zu dem monoklinoedrischen System mit dem Winkel Y (Zwischen der Vertical- und der Klinodiagonalaxe) = 900'0" gehoren." (Kokscharov's angle i is now referred to as 3). Kokscharov's axial ratios are (in modern form): a:b:c = 0.5774:1:1.6452 P = 90~0' They do not correspond with any known type, but, if c is doubled, the result; a:b:c = 0.5774:1:3.2904, compares favorably with the values for the 3-layer polymorph. These settings are of importance because they were adopted by Hintze (1897) for all the micas and by Baumhauer (1900) for muscovite. Goniometric measurements on muscovite have been made by Marginac (1847), Kokscharov (1875) Tschermak (1877), Baumhauer (1900), and Peacock and Ferguson (1943). Early measurements were often inaccurate and resulted in the proposal of several doubtful forms. Peacock and Ferguson (1943) regard the measurements of Baumhauer (1900) as superior and, after the transformation of his elements from the unsuitable settings of Kokscharov (1875), calculate a:b:c = 0.5764:1:2.2372 P = 95~39' which is considered representative of the geometrical elements of muscovite. In measurements of their own. Peacock and Ferguson (1943) have recorded the forms c,b,m,y,p,h,E, and o. The following table lists the goniometric measurements for the forms they found. 52

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN TABLE II PEACOCK AND FERGUSON (1943) MEASUREMENTS ON MUSCOVITE Measured Form Mea Calculated Mean cb = (001): (C ) 89 57 1/2' 90 000 cy = (001):(021) 77 13 77 20 1/2 cm = (001):(110) 85 02 85 06 ce = (001):(112) 61 45 61 53 co = (001):(111) 72 46 72 46 1/2 ch = (001):(112) 69 54 1/2 70 03 cp = (001):(ill) 82 13 82 06 1/2 Other forms have been established by other investigators but were not found on the crystals measured by Peacock and Ferguson (1943)5 They consist of (after transformation): n(130) reported by Marginac (1847) and Tschermak (1877) e(011) reported by Tschermak (1877) x(131) reported by Tschermak (1877) Calculated angles (by Peacock and Ferguson, 1943) for these forms are cn(001):(130) 87~ 10' ce(001):(011) 65 48 1/2 cx(OO1):( 31) 85 25 1/2 Doubtful forms are (032) (302) and (221) Peacock and Ferguson (1943, P. 75) illustrate a stereographic projection of the established forms for muscovite and also show a drawing of a crystal which consists only of the forms observed by them (Figso 35 4). Twinning in muscovite is common and is generally described according to the mica law. In this law (Ford, 1932) the twin plane is in the zone cM 55

Fig. 3. Stereographic projection of established forms for muscovite after Peacock and Ferguson (1943). ^r^-i ^ -^r^1010 m~ l l 130 o110 Fig. 4. Crystal drawing of muscovite forms found by Peacock and Ferguson (19435).

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 001A221 and normal to c(001), with (001) as the composition plane. Peacock and Ferguson (1943) state this law in a more general manner (p. 78): "The two individuals are in contact on a common plane c(001) and symmetrical by reflection in a plane (hht) which is perpendicular to c(001). Reflection in (hhf) gives a "right" twin which is distinct from a "left" twin resulting from reflection in (hh)." In the setting adopted by Hintze (1897), for example, the reflection plane is (110) which in the correct muscovite setting is (331). On the basis of the new-geometrical elements for muscovite, Peacock and Ferguson (1943) have shown that the twinned crystal is not symmetrical by reflection in the basal plane (001) for the calculated angle (001):(331) = 89~ 22 1/2.l Reflection in (331) would thus require an angle of 1~ 15' between the basal cleavages of the twin elements. This discrepancy may be avoided by defining the twin law of muscovite as twin axis [510, composition plane (001). The basal planes of the two individuals are thus brought into exact parallelism. Another type of twinning of more importance to this study may explain the occurrence of uniaxial (and small 2V) muscovites, such as that described by Axelrod and Grimaldi (1949), and lepidolites. It may be defined as twinning axis 0oj and composition plane (001) and is discussed in detail with reference to uniaxial lepidolite (page135). 3. Interferometric Data. The work of Tolansky (1948) has demonstrated that it is possible to obtain some information on various characteristics of mica by means of the application of multiple-beam interferometric techniques using Fizeau fringes and fringes of equal chromatic order. Data may be obtained regarding: (1) exceedingly minute topographic details of cleavage surfaces, including the direction of the cleavage step and the height of the step, (2) the thickness of the mica sheet studied, (3) the birefringence of the mica over a rather extended wave-length range, provided the refractive index is known for one wave length, (4) identification of various types of inclusions, among them solid inclusions of other minerals and vacuoles, and (5) variations in chemical compositiono There has also been an attempt to correlate the structure patterns as revealed by multiple-beam interferograms of doubly silvered mica sheets with electrical power factor and dielectric strength. Tolansky has demonstrated that muscovite from different localities, for example, Bihr, India muscovite as compared with Brazilian muscovite, gives different and typical interferometric patterns that are characteristic within a moderate range of variation. He has also attempted to show that these patterns differ among the various mica species: that muscovite patterns differ from those of phlogopite and biotite. The differences in pattern are due to variations in three features: (1) the 55

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN variation in average size of the area of uniform color tint, (2),the number of inclusions per cubic centimeter, and (3) the average size of the inclusiona The correlation between interferometric pattern and electrical properties is as yet apparently not complete or else the published results have not come to the writers' attention. However, it is suggested that a correlation may be expected between the dielectric power loss of the mica and the number and size of the inclusions, which, although they are often transparent because they consist either of liquid or gas or a combination of the two, are nevertheless strikingly revealed by the interferograms. Although this technique gives promise for a very exact grading of muscovite, it remains to be demonstrated, in our opinion, that the distribution, number, and size of these inclusions is relatively invariable in single sheets, in single books, or in books from single deposits, or else is variable within rather narrow limits, so that a technique that employs mica films of the order of 0.01 mm in thickness can be said to sample effectively either a single book or a deposit without requiring an inordinate number of samples. This technique has also apparently demonstrated that a few varieties of mica consistently can be cleaved to yield relatively large areas of very uniform tint, which according- to Tolansky, implies that the mica over this area has been cleaved true to a single molecular plane, not only on one side of the film, but on both sides of it. Such perfect mica sheets are of value. as windows or absorbers in radioactive experiments. One of the micas tested by Tolansky, which was from Tanganyika, showed a uniform area of 26 cm2. It would be very interesting to explore further the results obtained by multiple-beam interferometric techniques using a much larger number of samples from a great many different localities closely correlated with chemical analyses and optical studies. Within the past few years a rather considerable amount of study has been undertaken with respect to the screw-dislocation theory partly by means of interferometric methods. The theory and mechanism of growth by this method have been described in detail by Frank (1949). Dislocations have been reported by various investigators in single crystals of beryl, silicon carbide, paraffis and very recently in micas (Amelinckx, 1952A). Apparently dislocations may explain not only certain external features such as spirals, but also, according to its advocates, polymorphism in SiC and the micas. Briefly, the theory explains growth in natural crystals whereby the process of nucleation proceeds in such a manner that a spiral effect is produced; important factors in this process include height of the unit cell, supersaturation, and so forth. It appears quite probable that future work along these lines may yield important information concerning the cause of polymorphism. The results reported so far are quite convincing. 56

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN C. Optical Properties 1. Indices and 2Vo For normal muscovite the general ranges of indices of refraction are as follows: = 1.552-1,574, p = 1.582-1610, ~ = 1.587-1.616.'-o( = 0.036-0.049. However, values both below and above these ranges have been measured and reported. For example, values for a as low as 1.542 and 1.544 are recorded. Values for v as low as 1.579 and fort as low as 1.578 have been reported. 2V = 50~ to 47~, but some phengites have 2V that range from about 20~ to essentially 0~. All muscovite is (-); a A c = 0~ to -2~, P A a = 1~ to 3~,' = b. The optic plane is normal to (010) r>v weak to distinct. a. Variation of optical properties with composition: There have been several attempts to relate the variation in refractive indices and 2V to variations in chemical composition in the muscovites. Some of the more notable attempts are those of Gadeke (1938) who drew an essentially straight line showing a direct increase between values for P and weight percent.Fe203, embracing only six points. A more elaborate attempt has been made by Winchell, and the most recent results of his efforts are embodied in Fig. 254 of his textbook (Winchell and Winchell, 1951, p 368). Winchell correlates variation in optical and physical properties with percentages of three "end-member" molecules, muscovite, picrophengite and ferrimuscoviteo His diagram shows an essentially straight-line relationship between increasing P index of refraction and content of ferrimuscovite molecule. Another noteworthy attempt at correlation has been by Volk (1939). It is certain that the refractive indices increase markedly with relatively small increments of ferric iron, but it also seems likely from our studies that a lesser rate of increase is brought about by increasing ferrous iron. Thus the refractive indices of muscovite increase considerably with an increase in the Fe35Al ratio and also increase to a lesser degree with an increase in the Fe2:Mg ratio. Very likely the substitution of Mn for Mg also brings about a slight increase in the refractive indices. Many muscovites contain both ferrous and ferric iron, and the difference in the amounts of increase of the refractive indices caused by equal numbers of ferrous-iron atoms and ferric-iron atoms has not yet been established. We have made numerous plots of P values against weight percent FeO, against weight percent Fe203, against weight percent FeO + equivalent Fe203, and against weight percent A1203, and although all of these display a general trend, none shows a satisfactory direct relationship between increasing refractive indices and either increasing weight of the oxide or decreasing weight percent of the oxides, the latter in the case of A1203. For these plots we have employed as many as about 60 analyses, including those of Kunitz (1924), Jakob (1925B, 1929A, 1929C and Volk (1959), as well as some of our own and some by other investigators. One thing seems evident from these plots, and that is that there is a tremendous divergence in the quality in chemical analyses of muscovite. Possibly 57

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN also the results indicate that there is a considerable variation in the degree of accuracy of the measurement of refractive indices of muscovite. Probably the above relationships *hold for the birefringence which increases markedly with an increase in ferric iron and to a much lesser extent with increasing ferrous iron and manganese. The size of the optic angle decreases in a general way with an increase in magnesium, but also decreases in a general way with an increase in ferrous iron and possibly also ferric iron. Conversely, this relationship rnay be stated by saying that those micas which most closely approach the ideal composition of having little or no magnesium, ferrous iron, or ferric iron, are those with the largest optic angles. We have also plotted 2V and 2E values measured on micas analyzed by Volk (1939), Jacob (1925B, 1929A, 1929C), Kunitz (1924) and others, as well as micas analyzed specifically for this investigation, against weight percent Fe203, weight percent A1203 and weight percent Fe203 + FeO, and again, although general trends are apparent, the results are far from expressible as a straight-line function. It is interesting to note, however, that if one plots the 2E values obtained by one investigator, in this case those of Kunitz (1924) against the results of his analyses, a much closer approach to a straight-line function is obtained. This may again indicate a considerable difference in quality in both the chemical and optical results by different investigators. These general optical-chemical relationships are born out by the recent study of Heinrich and Levinson (1953) on rose muscovites. These muscovites approach very closely the ideal muscovite composition, containing generally less than 0.5 percent of combined Fe203, FeO, MnO, and MgO. They consistently have slightly higher 2V values than other muscovites and have consistently lower refractive indices than other muscovites. b. Changes of optical properties with temperature: Rinne (1925) has reported that muscovite becomes almost uniaxial when heated in air for about three hours at 1000~C. A second heating at 1000~C restores the original 2E. This transformation may be accompanied by structural changes, the character and extent of which are as yet unknown. We attempted to repeat this phenomenon but were unsuccessful. About two dozen runs were made under various conditions (i.e., in air and vacuum) in a high-temperature furnace capable of maintaining a temperature as high as 1200~C within 2 or 3 degrees indefinitely. The results indicated that at about 700~C the specimens (muscovite) change considerably in external physical appeal ance through the loss of color and through expansion. A decrease in 2V also accompanies the physical changes. It appears logical to assume that when water is driven out of the structure, or the iron is oxidized, the above changes take place. On cooling or reheating, the micas did not resume their original properties.

- ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN Others have studied various other physical and structural effects on micas heated to various temperatures under varying conditions. Powell and Griffiths (1937) noted that the thermal conductivity of muscovite varies little up to temperatures of 600~ (the maximum tested), whereas that of phlogopite decreased as much as 1/3 the initial value at 200~C. On cooling, the reversibility was incomplete. Wood (1937) showed, by means of X-rays, the interdependence of thermal and structural characteristics. Those phlogopites which exhibited a decrease in thermal conductivity at elevated temperatures, also showed displacement of the elemental units (layers) from the normal position at approximately the same temperature. Apparently the atomic arrangement was not affected. With the advent of differential thermal analysis many investigators have studied the micaceous and clay minerals (particularly the finegrained varieties) by means of this technique. Very little new information has been reported, however. 2. Variation of Color with Other Optical Properties and Composition. Muscovites show considerable variation in color, ranging from essentially colorless through shades of brown and buff (the so-called ruby- and rum-colored mica) to shades of green. The problem of color nomenclature has been studied in a systematic way by Jahns (1945) and by Jahns and Lancaster (1950). The chromophores in muscovite are Fe2, Fe3, Mn3, and Cr. It is possible that in a very few cases Ti may also play a very minor role. It has been demonstrated by Heinrich and Levinson (1955) that the rose muscovites owe their color to the essential absence of Fe2 and that Mn is dominant over Fe3. Under such conditions it is possible for Mn to be oxidized to the higher valence, in which state it becomes a powerful pigmenting agent, even though present in very small amounts. The dark-brown micas and those of reddish-brown tints apparently owe their color mainly to Fe3. This is particularly true for the very dark-brown micas which often have magnetite and hematite inclusions, probably due to exsolution. The green color of most muscovites is believed to be due to the predominating presence of Fe2. When either Fe2 or Fe3 is present in amounts of about 1 percent or more of the oxide, the pigmenting power of any Mh present is essentially masked. The colorless or nearly colorless muscovites apparently owe their lack of color to a close balance between the amounts of Fe2 + Fe3 and the amount of Mn. The coloring effect of chromium is well illustrated in fuchsites and mariposites, which owe their bright-green color to even relaxr tively minor amounts of chromium. Within a single pegmatite that contains both green and brown or ruby muscovites, the green variety consistently shows the larger 2E and will generally have lower indices of refraction than the brown variety. This is in accord with the relationship between increase in Fe5 with increasing indices and decreasing optical angle. In Table III the 168 analyzed muscovites have been classified into 2 main color groups: (1) green, olive green and greenish brown, and (2) light brown and ruby. Although the assignment for some specimens is somewhat arbitrary and the two color categories are undoubtedly somewhat 59

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN oversimplified, the color differences of the two are reflected in remarkably consistent chemical differences. Green muscovites are considerably richer in total Fe (expressed as Fe203) and in Mn, whereas ruby muscovites cont&in more Ti and usually more MgO. Mg is not considered to be a chromophore element. Cr, normally a green chromophore, actually is more common and abundant in the ruby muscovites. It is, of course, not possible, by spectrochemical analytical techniques, to analyze separately for Fe3 and Fe2. Green muscovites, although they contain more total iron, have generally lower refractive indices than ruby muscovites. The ruby muscovites owe their higher indices to their greater content of Fe3 and Ti, both of which cause much more marked optical changes than equivalent amounts of Fe2. TABLE III VARIATION OF COLOR WITH COMPOSITION IN MUSCOVITE (NEW DATA) For each district and composition, two numbers are given: the top number is the average percentage for green, olive-green, and greenishbrown muscovite; the bottom number is the average percentage for light-brown and ruby muscovite. Districts Avg, Avg, Avg, Avg, Avg, Districts Fe205 % MgO 2TiO2 A Mr) ACrg20 1. Franklin-Sylva district, 5.59 0.97 0.126 0.082 North Carolina and north 2.55 1.18 0.375 0.053 Georgia 2. Spruce Pine district, 4.35 1.21 0.111 0.056 N.C.; Shelby-Hickory 2.09 0.98 0.214 0.045 district, N.C.; ThamastonBarnesville district, Ga.; Alabama district 3. Average of micas in 3.88 1.07 0.120 0.072 districts 1 and 2 2.59 1.12 0.323 0.049 4. New England; Black Hills, 4.07 0.23 0.097 0.388 S.D.; Colorado; Petaca dis- 2.38 0.99 0.104 0.072 trict, N. Mex.; Montana; Latah Co., Idaho 5. Average of all muscovites 3.92 0.68 0.109 0.215 0.0007 2.46 1.09 0.281 0.056 0.0255 *Total Fe, expressed as Fe203 60

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN D. Zoning Color zoning is very common in muscovite, particularly core-margin types, with usually brown or ruby cores and greenish-brown or lighter-brown rims. Oscillatory zoning also occurs, usually with thin darker-brown zones alternating with broad light-brown bands, although less commonly green and brown zones alternate. Such zones are parallel with crystal outlines and thus are hexagonal or rhombohedral in outline. Some such zoned crystals also show concomitant. color variations in pairs of sectors which join at the center and probably indicate differential absorption during growth. Some ruby micas are characterized by oriented sets of closely spaced darker-colored streaks arranged at right angles in a cross hatch or grating pattern which resembles the gridiron twinning pattern of microcline. Cleavage sheets of such crystals also may show a division into three pairs of sectors, each pair being characterized by different orientation of the grating streaks. The streaks in adjacent sectors are oriented at 120~ to 60~ to one another. The sectors may result from twinning, and the grating pattern may represent embryonic exsolution. Color zoning may involve the c-axis direction with sheets at one end of the crystal differing in color from those at the other. Some muscovites that are marked by A structure show color streaks parallel with the reeves, somewhat different in shade from the rest of the sheet. Muscovite also commonly displays inclusion zoning of hematite, usually with stained cores and clear margins, although alternations of stained and clear zones also are recorded. Some examples of different types of zoning in muscovite are noted below. Minas Gerais, Brazil. "Symmetrically zonked color patterns, hexagonal in outline are sharply defined and most commonly are composed of recurrent zones of two different tints of ruby. Rarely, a rhombshaped ruby core is surrounded by hexagonally shaped zones of lightercolored ruby. In some specimens, cores of ruby mica are surrounded by zones of green and yellow-green mica. Three well-defined linear patterns are known. In one, linear bichromatic streaks are oriented along two sets of reeves; in another, called checkerboard or plaid, the bichromatic lines are at right angles to each other; the third, called'hen track'... has two sets of streaks parallel to the adjacent reeve sets, and the third is parallel to the limiting crystal face (010 or 110)." Two irregular color patterns have also been found in this district (Pecora, Klepper, and Larrabee, 1950). _______________________ 61 ___________

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Ejiba, Nigeria. Jacobson and Webb (1946) report three-dimensional color zoning in muscovite from an albitized pegmatite. At one end of a crystal the muscovite is apple-green and at the other it is lilac. The colors grade into each other along the length of the caxis. The only chemical difference in the two types is a slightly higher iron-manganese ratio in the former. San Diego County, California. Some zoned crystals have been reported from the Pala District, (Jahns and Wright, 1951) with green centers and intermediate zones that are yellow to white. Madison County, Montana. Stoll (1950) reports sheet mica that is colorless and shows a criss-cross pattern or banding of ram-colored areas. Taos County, New Mexico. Optical zoning was reported by Heinrich and Levinson (1953) in pink muscovite flakes from the Harding pegmatite. Many of the muscovite flakes have a very thin rim that displays lower birefringence than the main central part of the crystal. Rio Arriba County, New Mexico. Herinrich and Levinson (1953) also report small flakes of muscovite from the Petaca District that show pink and green colors grading into each other. Clear Creek County, Colorado. Muscovite from the Ajax pegmatite was reported by Hanley et al. (1950) to show alternating color strips of clear and ruby color. Southeastern Piedmont Area. Jahns (1945) summarizes the color characteristics of the muscovites of this area as follows: a) Many books are color zoned with gridiron or chessboard patterns, concentric color bands, strips, or faint irregular mottling. b) Nearly all books contain pale rims that are more greenish than the interior. This appears to be secondary zoning. c) Color variations in individual books are more pronounced where the mustovite is light- to olive-green. New England. Sterrett (1923) summarizes the types of zoning in New England muscovite as follows: "Some muspovite. shows variations in color that accord with crystal structure. The variations generally appear in bands 62

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN that follow the outline of the crystal. Thus,... one may see a dark rum-colored center surrounded by a fringe of light rumor yellow;... or the center may be light and the border zone dark. In some sheets there are alternating bands of varying color." Woodward (1951) reports an unusual type of structural zoning in the green muscovite of the Lord Hill pegmatite near Stoneham, Maine. One book of green muscovite grades from perfectly normal flat folia on one side of the specimen into a gray botryoidal folia on the other. The botryoidal structure may be due to late volume increase because of the addition of materials to the pre-existing muscovite. A group of selected specimens showing especially distinct color zoning was selected for optical study. Indices of refraction were measured on an Abbe refractometer and 2E was determined by the Mallard equation. The optical data are in Table IV: TA1BLE IV OPTICAL DATA ON SELECTED ZONED MUSCOVITES Intermediate Specimen Core Zone Margin No. 5. Gregory pegmatite, Jackson Co., N. C. Color green clear a 1.562 1.565 P03B ~ 1.598 1.602 1.603 1.612 2E 72 67~20' No. 7. Jasper pegmatite, Jackson Co., N. C. Color dark-brown medium-brown light-brown ca 1565 1.564 1.566 B 1.602 1.601 1.601 1.606 1.605 1.605 2E 67025t 65015' 62040' 63

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN TABLE IV CONT. Specimen Core Intermediate Margin Zone No. 9. New Wolff pegmatite, Franklin-Sylva District, N. C. Color clear light brown cIa 1.569 1572 p3 ~.1.607 1.612 g.1.613 1.-615 2E 54~0 lo 59~15' No. 15. Sheep Mountain pegmatite, Franklin-Sylva District, N. C. Color stained clear ca 1.566 1.567 1.602 1.604 1.608 1.608 2E 61~40o 60~10l No. 46. Ledford Cove pegmatite Franklin-Sylva District, N. C. Color stained clear a 1.562 1.565 P 1.598 1.601 1(g~ 1.607 1.607 2E 62030v 63030' No. 57. Poll Miller pegmatite, Franklin-Sylva District, N. C. Color medium-brown clear a 1.560 1.560 1.595 1.593 1.604 1.607 2E 70~00' 69045' No. 64. Slagle-prake pegmatite, Franklin-Sylva District, N. C. Color clear spotted a 1.564 1.559 1.601 1.596 1.607 1.606 2E 65~3050 63~ 4 - -- -— 64

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN TABIE IV CONT. Specimen Core Intermediate Margin Zone No. 75. Big Ridge pegmatite, Franklin-Sylva District, N. C. Color light-brown medium-brown a 1.562 1.566 P 1.599 1.601 1.606 1.606 2E 65030' 65~40 No. 82. Big Ridge pegmatite, Franklin-Sylva District, N. C. Color light-brown medium -brown a 1.565 1.566 P 1.599 1.603 6 1.605 1.606 2E 64~50' 64~50' No. 124. Plumtree pegmatite, Spruce Pine District, N. C. Color med ium- green light-green a 1.570 1.569 ePB~ 1.610 1.603 1.615 1.609 2E 64015 67~10' No. 128. Wildcat pegmatitej Spruce Pine District, N. C, Color dark-green light-green a 1.570 1.569 pfB ~ 1.610 1.606 1.616 1.616 2E 70000' 67030'

ENGINEERING RESEARCH INSTITUTE *UNIVERSITY OF MICHIGAN TABLE IV CONCL. Specimen ntermediate Margin Specimen Core Margin Zone No. 190. Muscovite pegmatitey Latah County, Idaho Color light-brown clear a 1.562 1.562 P 1.598 1.599 1.602 1.604 2E 63015' 635 No. 374. Hebron, Maine Color medium-tan light-tan clear a -- 1.563 1.560 P 1.594 1.598 1.591 1.603 1.601 1.598 2E 67010? 650~15 61050' The above data indicate that slight differences in indices of refrac tion and in size of the optic angle accompany variations in color. The darkercolored zones and those that'are stained or contain inclusions tend to have higher indices and larger optic angles. II. MINERALOGY OF VARIETAL MUSCOVITES A. Oellacherite and Other Reported Barium-containing Micas. The term oellacherite was first used by Dana (1867) for the bariumbearing mica (4.65-5.91% BaO) from Pfitschthal, Tyrol, originally called margarite by Oellacher (Kenngott, 1860). Chemical analyses of oellacherite are reported by Rammelsberg (1862, 1879) and Sandberger (1875). The occur:rence has been described by Weinschenk (1896, p. 471). A specimen of oellacherite (No. 1092) originally from the Karabacek collection, from the type locality, was borrowed from the Harvard Mineralogical Musuem. The oellacherite is in flakes 1 to 2 mm across, and is pseudomorphous after a curved, bladed mineral, probably kyanite, in white quartz. X-ray studies show that it has crystallized with the normal muscovite structure. A flame test for Ba gave completely negative results. Since the analyses are old, some question arises regarding their validity. Spectrochemical analyses of about 200 muscovites, mainly from pegrnatites, indicate 66

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN that the BaO content of such micas ranges from 0.0002 to 1.1 percent. Simpson (1932, p. 67) reports a muscovite from Lower Chittering, S. W., Western Australia, pseudomorphous after corundum; this mica is reported to contain 1.16 percent BaO. The mineral from the type locality has the following optical properties: = 1.576 = 1.578 2 = 43~ These values are well within the lower part of the range for normal muscovite. It is to be expected that if significant amounts of Ba are present, the indices should be somewhat higher than those normal for muscovite. Kultiassov and Dubinkina (1946) have described a vanadium-bearing variety of oellacherite from the Karatau, U.S.S.R., vanadium deposits. Three varieties of the mineral are distinguished1. Light-green, silky, fibrous aggregates, intimately intergrown with columnar quartz. 2. Dark-green, pearly tablets imbedded in coarse quartz veins. 3. Dark grayish-green flakes and scales on fault surfaces. Chemical analyses of the three varieties show similarities to oellacherite with respect to the BaO content and to roscoelite with respect to the V203 content: 1. 2. 3. V203 6.18 17.92 16.44 BaO 7-15 5.27 7.20 X-ray powder data are presented for the fibrous and flaky varieties (1 and 3). Inspection of the d-spacings reported for the fibrous type indicates the presence of considerable quartz. Practically all the spacings (and comparable intensities) listed on the ASTM card for quartz are recorded. When these lines are subtracted from the list of d-spacings, only weak reflections rema No mineral with these remaining spacings could be found on the ASTM cards. In many of our samples we have noted that a relatively small amount of quartz contamination will produce marked effects on mica powder patterns. These observations cast considerable doubt on the homogeneity of the material on which the analysis was made. The reported spacings for type 3 are equally confusing. Although the first line recorded has d = 9.955, which is close to that of an important muscovite line, quartz lines also predominate, and the remaining spacings are unassignable. They do not correspond, however, to the muscovite or even thegeneral mica spacings. If the analyzed material represents the same sort ofmixture of which the X-ray powder photograph was taken, then its value is nil. The authors (Kultiassov and Dubinkina, 1946, p. 190-191) admit that (in translation) "The content of Si02 in the analyses appears to be somewhat high. 67

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN This is because the mineral forms fine intergrowths with quartz, and as a result it was impossible to get completely pure material for analysis." They also suggest that the CaO and S03 reported may be due to admixed gypsum. We have compared the lines remaining after quartz lines were subtracted with the powder diffraction pattern of gypsum and find no correlation. There is no record that the analyzed minerals were first examined microscopically for purity. In view of this information and in the light of the reinterpretation of the X-ray powder data, the identity of this material can hardly be considered well established. Through the courtesy of Dr. S. 0. Agrell, we have been permitted to examine a hithertofore undescribed specimen (No. 1349) of a "titan-oellacherite" from a slag. X-ray studies of this material by means of the Weissenberg method indicate that the mica has crystallized as the 1-layer monoclinic polymorph. Because of the high MgO content (20.19 percent) it seems undesirable to call this mica oellacherite; barian phlogopite appears more acceptable. Furthermore, many phlogopites have 1-layer form, whereas muscovite does not. The formula of this material is: (K0.8, Ba0.8Ca0.4NaO.) (Mg4.5Fe20.l1Mn)Ti0o.3Fe5) (Si4.1, A13.3, Tio.6) 021.0 (F2.4, O0H.6). A barian biotite (barytbiotite) with 5.11 percent BaO has been reported by Daub (1913) from Kaiserstuhl, Germany, but the presence of 8.17 percent CaO in the analysis makes its validity suspect. Many analyses of phlogopites with from 0.4 to 1.3 percent BaO have been reported, and, in general, Ba seems to occur more commonly and in larger amounts in phlogopites than in any of the other micas. A synthetic Ba mica (13.84 percent BaO) has been described by Daimon (1952); it is also a barian phlogopite. The name "oellacherite" should be discarded and "barian muscovite" should be substituted, if it can be shown that the original material actually contains the amount of BaO reported in the old analysis. Barium-muscovite from Franklin, N. J.: A mineral with 9.89 percent BaO, from Franklin, New Jersey, described by Bauer and Berman (1933), has been called barium-muscovite. Specimens of this material (Nos. 540, 1095, 1127) have been obtained from the Harvard Mineralogical Museum and from the U. S. National Museum. All specimens are massive and pink in color, not unlike some so-called oncosines. The formula, as calculated from the Bauer analysis, is: (K1.2 Ba0.6 Na0.4, Cao.l) (MgO.3Zn0.2MnO1)A1.4 (Si5.9 A121) 020,1 0H..9 which is in close agreement with the requirements of the formula for a muscovite-type mica. Microscope examination demonstrates the essential homogeneity 68

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN of the material. It consists of exceedingly minute flakes of a colorless mineral, finely intergrown. The mean index of refraction is about 1.59; the birefringence is that of fine-grained muscovite. Examination by means of the electron microscope also substantiates the micaceous nature of the mineral. X-ray powder photographs taken of three specimens yielded nearly identical patterns with small differences in intensities of a few lines between the material from the Harvard Collection and that from the National lMuseum. Although it is of the general mica type, the pattern is unique and is unlike that of any known mica polymorph. Table 54, column 3, reports the d-spacings for this mica and compares them with those of normal muscovite (Column 1, Table 54. The above observations indicate that the material is homogeneous and that it is a mica, but that it can hardly be referred to as a bariummuscovite. Until the exact structure has been determined, the position of the mineral in the mica group remains uncertain. Since its structure is appaently unique, it should perhaps be accorded species rather than varietal status and should receive a name of its own. In conclusion, it may be said that the status of the barium-containing micas is confused. It appears that a barian muscovite probably does not contain more than 5.91 percent BaO, if the old analyses are reliable. The effect of Ba on the structure remains to be clarified. B. Ferrian and Manganian Muscovite A survey of the literature reveals that Fe+3 has been reported up to a maximum of 5.70 percent Fe203 in muscovite (Tschermak 1878). Ferric, as well as ferrous iron, substitutes for octahedral aluminum. Many examples of muscovite with 2 to 4 percent Fe203 are available in the literature. On the basis of new data obtained by spectrochemical studies obtained for this report, the above observations are substantiated. Rose muscovites and other late hydrothermal micas have the least amount of ferric and ferrous iron, usually less than 0.5 percent. Several high-ferrous muscovites have been reported such as that by Wulfing (1886) (FeO = 6.55 percent). However, highferrous muscovites are usually associated with phengites; the rise in Fe+2 or Mg, or Mg + Fe2, as a generalization, accompanies an increase in Si. The structure of ferrian muscovite is most probably 2M muscovite. This statement is based on X-ray studies of the "alurgite" from Cajon Pass described by Webb (1939) which has been shown by Hendricks and Jefferson (1939), and verified by us, to have the normal muscovite structure; it is actually a ferrian (5-32 percent Fe203) muscovite. Mangan-muacovite: Eskola (1914), in describing the pegmatites of southwestern Finland, notes (p. 37)o "A particular manganese-mineral, an intensely blue-colouredfinecrystalline manganese-muscovite containing 2.30% MnO, was found at Mattkarr." 69

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN This or similar material was studied spectrochemically by Erametsa, Sahama and Kanula (1943), who determined the presence of 1 percent Rb20 and 0.03 percent Cs20. Pehrman (1945) also describes a blue "radial" mica from Mattkarr, which probably represents similar material. He found it to have 2V = 39~20' by means of the universal stage. A specimen of the Eskola mineral was borrowed from Harvard Mineralogical Museum (No. 87441) and studied. The material is deep purplish-blue in color, very fine-grained, and shows no megascopic structural features. Microscopic examination reveals that is is a mixture of four different species, in order of abundance: 1. Shreds of colorless, fine-grained mica, n = 1.56 2. Shreds of pale lilac to purplish-gray mica, n = 1.59 5. Quartz 4. Scattered fine grains of a mineral with low birefringence, numerous minute inclusions, and n = ca. 1.63; probably apatite. The purplish mica, which has the birefringence of normal muscovite, is subordinate to the colorless mica and occurs intergrown with it as scattered clumps of irregular, interwoven flakes, which appear to have been formed at the expense of the colorless mica. Studies of the purple mineral by means of the electron microscope reveal that it is distinctly micaceous. The powder X-ray photograph of this material also shows the presence of very strong quartz lines. If these are subtracted from the pattern, the remaining lines are probably indicative of a mica, because of the presence of a strong reflection near lOA; but the structural type is unidentifiable. In view of the optical demonstration that two intergrown micas are present, the pattern may also represent a composite of two polymorphs. From the finegrained and intimate nature of the two-mica intergrowth and the similarity in their physical properties, it is unlikely that separation of the two can be effected by ordinary mechanical means. A spectrochemical analysis (spec. 1093) of the Harvard material labeled mangan-muscovite is given in Table V. This analysis corresponds with one to be expected from a mica or a mixture of micas contaminated by quartz (high SiO2) and by apatite (1.05 percent CaO). It can be stated the purple mineral is a mica, probably of the muscovite type, but it is to be noted that a number of other muscovites with larger amounts of MnO have been reported, which have not been designated mangan-muscovite. The varietal name "manganian muscovite" is preferable, in any event. Although only a maximum of 2.32 percent MnO has been reported in muscovite (Ellsworth, 1952) the writers believe that the manganian-muscovite variety is justified. Most muscovites contain small amounts (< 1 percent) 70

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN MnO which undoubtedly occupy octahedral positions. These micas are normal with respect to optical, chemical and structural properties. The maximum amount of manganese that can enter the muscovite structure, or its effects, is as yet unknown. TABLE V SPECTROCHEMICAL ANALYSIS OF MANGAN-MUSCOVITE (ESKOLA, 1914) Si02 - 49. A1203 - 32. K20 - 10. Fe203 - 0.15 MgO - 0.21 Na2O - 0.08 MnO - 1.6 BaO - 0.0008 Ga203 - 0.012 SrO - 0.004 CaO - 1.05 Analyst: Charles E. Harvey C. Chromian muscovite Until 1946, the chrome micas were in a state of considerable confusion. At that time, Whitmore, Berry, and Hawley reviewed the whole problem and clarified many outstanding inconsistencies. Fuchsite was first described by Schafhautl (1842) for the green chrome mica from Schwarzenstein in the Zillerthal. An analysis of this material is reported in Dana (1882) as containing over 5 percent Cr203. Six additional analyses are presented in Dana (1892) on chrome micas from Tyrol; Sysevsk; Urals; Belgium; Maryland; Aird Is., Lake Huron; and Ouro Preto; in these analyses Cr203 is reported from 0.84 to 3.95 percent. Additional analyses on fuchsite from Switzerland have been reported by Prior (1908) with 0.87 percent Cr203; from South Africa by Partridge (1937) with 0.85 percent Cr203; from New Zealand by Hutton (1940) with 0.27 percent Cr20O and (1942) with 4.60 percent Cr203; from Colorado by Wherry (1915) with 6.08 percent Cr203 (the highest recorded chromium content of fuchsite); and from Pointe du Bois, Manitoba, by Whitmore, et al. (1946) with 4.81 percent Cr203. Additional occurrences of material called fuchsite have been reported in the 71

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN literature. Typical of recent interest is the occurrence in the Colorado Plateau as discussed by Kerr, et al. (1951). Partridge (1937) preferred the term'chrome-muscovite" to "fuchsite." Odman (1938) favored "mariposite" for a low chromium mica even though his material is optically similar to low chromium fuchsites. Hutton (1942) suggested that: (1) the term "mariposite" be abandoned, (2) potash micas with less than 1 percent Cr203 be described as chromiferous muscovite, and (3) micas with more than 1 percent Cr203 be described as fuchsite. Mariposite was named originally for the green mica occurring in the Mother Lode district in California. Analyses of this mariposite and one from Ontario are tabulated by Whitmore, et al. (1946). The feature that distinguishes mariposite from fuchsite is the high SiO2:A1203 ratio and consistently low (0 to 0.78)percent Cr203 content. This has been pointed out by Schaller (1916, 1950), Winchell (1925), and Whitmore, et al. (1946). In agreement with Whitmore, et al. (1946), mariposite is considered as chromian phengite. Very little had been done in the way of X-ray studies of the chrome micas before Whitmore, et al. (1946). Only Pauling (1930) made observations on fuchsite as an outgrowth of his study of muscovite. Whitmore, et al. (1946), on the basis of X-ray studies on material from Pointe du Bois, Tyrol, and Ontario concluded that fuchsite and muscovite are structurally identical. A new chemical analysis on fuchsite from Pointe de Bois yielded the following formula: (K,Na,Ca)3.64 (Al,Mg,Cr,Fe3,Fe2)7.92 (A14,Sil2)038.61(OH)9.39 which compares favorably with the formula for muscovite. They further concluded that in fuchsite Cr may replace Al up to 4.81 percent Cr203 which is close to one atom per unit cell. On the basis of these new data, the term "chromian muscovite" is recommended. X-ray studies in this laboratory by means of the Weissenberg method were carried out on about one dozen fuchsites including the material from Pointe du Bois described by Whitmore, et al. (1946)and the chromium-containing muscovite described by Hutton (1940). The results obtained are in complete agreement with those reached by Whitmore, et al. (1946); the term "chromian muscovite" is thus adopted. Mariposite is considered by us (in close agreement with others) to be a variety of phengite and as such is discussed under that heading. The optical properties of fuchsite are, likewise, discussed by Whitmore, et al. (1946). They show the following variations: a = 1.559 to 1.569, A = 1.593 to 1.604, If = 1.597 to 1.611, Y - = 0.035-0.042, and 2V=530~ to 46~. The average p = 1.598 and 2V = 36~. The optic plane is 72

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN normal to (010). X<Y<Z; pleochroism has been noted by some authors as very distinctive, with a = robbin' s-egg blue 3 = yellow, yellowish-green 4 = blue green In thick flakes, dispersion is noted as distinct with r > v. Whitmore, et al. (1946) conclude that the indices of chromian muscovite (fuchsite) increase with increasing chromium. Some of the indices overlap those of mariposite but are generally lower. 2V on mariposite has been reported as small (0~ to 14~). The chrome micas variously classed as chromiferous muscovite, fuchsite and mariposite occur chiefly in two types of geological environments: (1) hydrothermal veins and replacement deposits and (2) metamorphic schists, gneisses and quartzites. The first type of geological environment has been subdivided by Whitmore, et al. (1946) into three variations. (1) with ankerite, quartz sulfides, and gold, (2) <with biotite and actinolite, and (3): with cordundum, biotiteand kyanite. These investigators conclude(p., 15), that the chrome micas "have invariably been formed either by metasomatic processes or deposited directly with other minerals and veins. Minerals associated with them suggest they were formed under conditions ranging from moderate to fairly high temperatures and pressures...." It may very well be that many of the chrome muscovites, if not most of them, have been formed by deposition from hydrothermal solutions or by metasomatic replacement as these investigators suggest. However, there are numerous examples of the occurrences of chrome micas in regional metamorphic rocks, particularly schists and quartzites in which there is no evidence of metasomatism. Chromiferous muscovite quartzites have been encountered by the writer in several districts in southwestern Montana. The rocks are fine-grained, well foliated quartzites whose color varies from very pale green to deep blue green. The depth of the color is directly proportional to the amount of chromiferous muscovite present. No peridotites lie close to the bands and lenses of green quartz, nor is there any evidence of hydrothermal alteration of these rocks. Fuchsite also occurs in some marbles as is evidenced by the occurrence at Marble, Colorado, (Wherry, 1915) and the occurrences at Binenthel, Switzerland, in dolomites (Prior, 1908). Several aberrant occurrences of fuchsite are recorded by Taddei (1940). He notes fuchsite in a pegmatite in Tessin, Switzerland. Fuchsite also is reported to occur in the emerald-bearing veins along with calcite, apatite, quartz, pyrite and albite at Somondoco, Colombia. 73

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Thus, although it appears that these additional recorded appearances of fuchsite substantiate the general conclusions of Whitmore, et al. (1946) that most of these micas are of hydrothermal or metasomatic origin, there are apparently also a few cases in which the chromiferous muscovite formed under regional metamorphism without assistance of hydrothermal solutions. D. Phengite 1. Introduction. The term "phengite" includes several kinds of high-silica muscovites and it is therefore necessary to discuss the varieties individually. (a) Phengite E2 (MgFeA1)4(Si7A1)020 (H)4 -- related to normal muscovite but with high silica content and increased Mg and Fe2 in place of Al. The effect of this substitution on the optical constants is discussed in detail on page 57. Two optic phases: biaxial and uniaxial (hexagonal phengite). (b) Mariposite (preferably, chromian phengite and hexagonal chromian phengite). Chemically, mariposite has decidedly lower Cr203 (0.10 to 0.78 percent) and A1205 contents than fuchsite but decidedly higher SiO2; the latter feature necessitates it being considered as a variety of phengite. The Al:Si ratio for atoms in four-fold coordination varies between 2:14 and 1:15 as compared with 4:12 in fuchsite. The structural formula of mariposite as calculated by Whitmore, et al. (1946) compares favorably, nevertheless, with fuchsite and muscovite. The indices of mariposite, on the basis of the available information given by Whitmore, et al. (1946) are: a = 1.56 to 1.58; p = 1.601 to 1.624; X = 1.60 to 1.63. 2V is usually small but one is reported as much as 40~. The indices of mariposite are generally higher than those for fuchsite of similar Cr203, and in some cases even higher than for Cr203-rich fuchsites. Some mariposite shows pleochroism X = deep blue green and Y = Z paler green. Overlapping of value of the optical constants for mariposite and fuchsite make accurate determination difficult by these methods. (c) Alurgite —for a discussion of alurgite see pp. 76 (d) Leucophyllite described by Starkl (1883) from Austria, is an exceedingly high-silica muscovite (57.11 percent Si02). The 74

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN analysis, however, shows only 3.39 percent K20 (plus 1.42 percent Na20) which may indicate that the analyzed material was impure, leached, etc. Schaller (1950) has discussed the series muscovite-leucophyllite, in which he includes phengite, alurgite and mariposite. He uses the formula IKAMg(Si)40l0(0H)21 for the leucophyllite end-member; this varies considerably from that reported by Starkll (1883) with regard to alkalies and aluminum. Schaller (1950), in interpreting the chemical composition of the high-silica sericites, places them into a muscovite-leucophyllite series and demonstrates that they also contain important amounts of a divalent element, usually magnesium. He also notes that their optical properties and specific gravities cannot be correlated with variation in chemical composition. Varieties placed by Schaller (1950) in this series include phengite, mariposite and alurgite (in part). Other investigators including Whitmore, Berry, and Hawley (1946), Hutton (1940), and Webb (1939) also have discussed the relationships of these micas to each other and to other members of the mica family. However, very little structural work has been done on them. Hendricks and Jefferson (1939) note that uniaxial alurgite from St. Marcel, Italy, has crystallized as the 3-layer hexagonal polymorph, whereas the so-called alurgite from Cajon Pass, California (Webb, 1939) has the normal muscovite structure. Whitmore, et al. (1946), in their work on the chrome micas, note that mariposite conforms chemically with phengite, which agrees closely with Schaller (1950), and they regard it as a chromiferous variety of phengite. However they had no specimens of mariposite suitable for single-crystal X-ray studies and state that none are recorded by other investigators. High-silica micas commonly have been reported to exhibit two optical phases, a biaxial and a uniaxial or nearly uniaxial phase, which may be intergrown in the same sheet. These two optical phases are structurally distinct. The uniaxial type has a 3-layer hexagonal structure whereas the biaxial type has the normal 2-layer monoclinic muscovite structure. Structural determinations can be made either by means of Weissenberg photographs, as demonstrated by Hendricks and Jefferson (1939), or by means of powder photographs, as shown by Levinson (1953). We have obtained a number of high-silica muscovites suitable for X-ray investigation. This section correlates structural types with optical variations in these micas and presents our conclusions on their classification. 75

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Dr. H. S. Yoder (personal: communication) has informed us that the high-silica sericites from Amelia, Va. described by Glass (1935) have the 1-layer monoclinic structure. We have not yet had the opportunity to study these micas. It should be noted that these micas are of pegmatitic origin, whereas the high-silica micas discussed in this section are generally of different paragenesis. 2. Alurgite. Alurgite, St. Marcel, Italy: Both uniaxial and biaxial alurgites from St. Marcel, Italy, the type locality, were X-rayed. The uniaxial sections have crystallized As the 3-layer hexagonal polymorph and the biaxial parts as the normal 2-layer muscovite polymorph. Penfield (1893) noted that 2E was as large as 57~; Larsen and Bierman (1934, p. 165) report alurgite, presumably from the same locality, with 2E as much as 98~. The Penfield (1893) analysis, the only analysis of the type alurgite, apparently was made on material of both uniaxial and biaxial character. Penfield (1893, p. 289) states: "The largest cleavage plates that have been observed are about 15 mm in diameter. When examined in convergent polarized light some show a uniaxial, others a biaxial interference figure, the double refraction being rather strong and negative.... The uniaxial character is undoubtedly the result of twinning, for plates can be found, which in some parts are uniaxial, in others biaxial, and where the plane of the optical axes is in two different positions. Some, moreover, give a confused interference figure, as would be expected from two or more biaxial plates, placed one above the other in twin positions. There is no difference in specific gravity between the uniaxial and biaxial plates, they are identical in behavior before the blowpipe and there is every reason for believing that they are the same chemical substance." Although it appears possible to simulate a uniaxial interference figure by systematically stacking thin muscovite sheets (Reusch, 1869 and Sugi, 1940), it is structurally impossible to arrange sheets with the 2-layer muscovite structure in any way so that the 3-layer hexagonal structure results. Although Penfield (1893) states that alurgite is monoclinic, he notes (p. 289 ) "When treated with a mixture of strong hydrofluoric and sulphuric acids the biaxial plates show monosymmetric etching figures while those of the uniaxial plates are triangular or hexagonal, similar to those of muscovite and biotite, described by H. Baumhauer." The 3-layer hexagonal structure on St. Marcel alurgite with 2V = 0~ was first established by Hendricks and Jefferson (1939). 76

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Alurgite also is reported to occur at Ceres, Valli di Lanzo, Italy (Zambonini, 1922; Gennaro, 1925). AlurgiteJ,East Tyrol, Austria: Meixner (1939) reports alurgites from several localities in East Tyrol, Austria. He confirms the presence of uniaxial and biaxial phases in the St. Marcel mica and states that the red mica from "Matrei", East Tyrol, has (p. 697), "Genau die gleichen Eigenschaften... " Three specimens of alurgite from this locality were obtained from the Museum of Natural History of Vienna. Our examinations show: Vienna General Optical Number Appearance Characteristics Structure J2407 Very minute 2V = ca. 35~ 2-layer monorose flakes clinic muscovite (Weissenberg). G5930o Cryptocrystalline, Birefringent, 2-layer monodeep red. Assoc. 2V not determinable clinic muscoMn-oxides vite (powder) G5931 Cryptocrystalline, Birefringent, 2-layer monodeep red 2V not determinable clinic muscovite (powder) Since chemical analyses of material from this locality are not available,, the mineral cannot be classed as alurgite with certainty. The existence of two optical phases whose presence is intimated by Meixner (see above) was not substantiated. Alurgite, India: Fermor (1909), in a systematic study of the manganese deposits of India, reported four occurrences of red and pink mica which he provisionally regarded as alurgite. A small specimen labeled "Alurgite, Ikabis State, Central India", was received from the Harvard Mineralogical Museum. It bears a general color resemblance to alurgite from St. Marcel; the fragment is biaxial with a large 2V, Weissenberg photographs indicate that it has the normal 2-layer muscovite structure. Alurgite, Cajon Pass, California: Type material of the alurgite described by Webb (1939) was rechecked by us and the results are identical with those of Hendricks and Jefferson (1939). This mica is not an alurgite, chemically, as has been pointed out by Schaller (1950) (also by Webb, 1939), 77

-ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN for its silica content is that of a normal muscovite. Since its structure is that of normal muscovite too, it must be classed merely as a ferrian muscovite. Professor Webb (personal communication) states that he is no longer convinced that his material is true alurgite. Alurgite, Ultevis, Sweden: Alurgite described by Odman (1950) contains only 43.67 percent SiO2 and 4.36 percent Fe203 and is chemically, therefore, a normal ferrian muscovite. X-ray photographs of flakes of this mica show that it has the normal 2-layer muscovite structure. We believe that the term "alurgite" should be dropped entirely and material of such character be referred to as manganian phengite (or hexagonal manganian phengite). For the historical treatment, we shall temporarily continue to use the term alurgite. 3. Mariposite. No mariposite suitable for single-crystal work could be obtained. However, two specimens of fine-grained mariposite, one from Carson Hill, Calaveras County, California and the other from San Francis= quito Canyon, Los Angeles County, California, described by Murdoch and Webb (1938), have been studied by the powder method. These specimens possess the normal 2-layer muscovite structure. Unfortunately, the material is so fine grained and has such poor crystal development that it is impossible to determine the optical properties with certainty. Although Murdoch and Webb (1938) state that the mariposite is found in flakes 10 to 12 mm in diameter, these flakes actually are composites of innumerable, much finer-grained flakes. However, they show considerable birefringence on basal cleavages, indicating that they are biaxial. Specimens of mariposite with a small NV, described by Maopf (1929), from the Mother Lode area of California could not be obtained. Sandrea (1950) described a mariposite with 2V<100 from Hoggar, Algeria, but again samples were unobtainable Whitmore, et al. (1946) mention that specimens from the Rawhide Mine, California, (Harvard No, 86911) have small 2V's. These specimens were obtained from Harvard for restudy but we are unable to substantiate the presence of mariposite flakes with small optic angles. The mariposite is extremely fine-grained, and the few shreds that were isolated show considerable birefringence. All efforts to isolate enough material for an X-ray powder photograph wer-e likewise unsuccessful; the photographs show the presence of quartz and a few extraneous lines. The table of optical properties of mariposite given by Whitmore, et al. (1946, p. 14) leaves the impression that a small optic angle is very common in mariposite. Whitmore, et al. (1946) did not study any of the mariposites by means of X-rays. We predict that any uniaxial or nearly uniaxial mariposite that may be found. will be shown to have crystallized as the 3-layer hexagonal polymorph, for mariposites are merely chromian high-silica muscovites e ---------- 7 8 —-----

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 4. Phengite. Phengites from Italy: The phengite analyzed and described by Pagliani (1937) from Fontane, Germanasca Valley, is reported to have a 2V of about 11~. The 2V of the type material we received is larger than 11~, ranging generally from 18' to 356 (estimated). Weissenberg photographs of sections with 2V near 350 show the normal 2-layered muscovite struc ture, but photographs of the sectiOns with the smaller 2V have combined patterns of what appears to be the 2-layered muscovite form and the 3-layered hexagonal polymorph. It is probable that patterns of the phengite with 2V = 11~ might be those of the 3-layered hexagonal polymorph alone. Specimens of a mica called muscovite by Jakob (1929B, No. 43) have been received and studied. This mica, from Passo di Ganano, Soazza, Val Mesocco is uniaxial and has crystallized with the 3-layer hexagonal structure. It is a high-silica muscovite (Si02 = 49.01 percent), and has the required, relatively large content of a divalent element (MgO = 3.91 percent). This mica was called muscovite by Jakob (1929B), rather than phengite, probably because of its occurrence and paragenesis, for the term phengite apparently has been used by some investigators for high-silica muscovites of secondary origin. Although this mica is predominantly uniaxial, several smal areas in a few flakes show birefringence and have a moderate 2V. One such isolated biaxial part proved to have the normal 2-layer muscovite structure. Phengites with the two-layer muscovite structure: Jakob has analysed five other high-silica muscovites (Table VI) all of which we have checked and found to have the normal 2-layer monoclinic muscovite structure. TABLE VI: PHENGITES WITH THE TWO-LAYER STRUCTURE ANALYZED BY JAKOB (1925B,1929B, 1929C) Jakob's No. Jakob's Designation SiO2 MgO FeO Fe203 -. "'- i -,,''..,:',''' ~:,... 18 muscovite 50.79 2.80 1.34 2.94 44 phengite 50.64 4.15 1.01 2.97 45 sericite 48.80 4~74 0.00 5.86 46 sericite 50.20 3.82 0.00 9.62 54 muscovite 50.04 3.64 0.85 2.44 Jakob's (1929B) Nos. 45 and 46, although they have a mica composi~t tion and the muscovite pattern as shown by powder photographs, are actually fragments of sericite schist. It was thought that the high silica content of these specimens might be due to admixed quartz. Microscopic examination revealed about 3 to 5 percent admixed epidote in No. 45, but No. 46 is essentially pure mica. 79

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN There exist, of course, many additional examples of muscovites described by other investigators, which are silica-rich and probably have only a biaxial, two-layer phase. The examples cited above suffice to demonstrate that not nearly all high-silica muscovites, of various varieties, have small optic angles and the 3-layer hexagonal structure. Many phengites, probably even a majority of them, are biaxial and possess the normal 2-layer monoclinic muscovite structure. 5. Muscovite, Sultan Basin, Washington. The only reported muscovite with a small 2V and without unusually high Si + Mg and/or Fe2 is the 3-layer monoclinic muscovite described by Axelrod and Grimaldi (1949). A portion of their analyzed material was made available to us for restudy by means of Weissenberg photographs. As explained on page 47, a study of our Weissenberg photographs substantiates the existence of the 3-layer polymorph but leaves the system in doubt, for we do not consider the differences in diffuse scattering to be of sufficient magnitude to warrant classing the structure as monoclinic. If this mica should be considered as being 3-layer hexagonal and if the SiO2 determination is correct, it will be the only muscovite with small 2V studied by us in which the three-layer structure is not associated with high silica content. 6. Other reported uniaxial or near-uniaxial muscovites. The mica from Mt. Mucrone, Switzerland, described by Kunitz (1924) is uniaxial and is called phengite (SiO2 = 50.45 percent) by him. Specimens of this material could not be secured for X-ray studies. Axelrod and Grimaldi (1949, p. 560) note, "'Carl Schmidt found, in a gneiss from the Adula Mountains, Switzerland, a green pleochroic muscovite that was uniaxial. It became biaxial on heating and on cooling became uniaxial again." Analyses of what appears to be similar uniaxial material, and also of an associated biaxial mica (2E = 52~) from the Adula Mountains, alsc collected by Schmidt, have been reported by Wulfing (1886). Mica I Mica II Si02 47.69 47.72 Ti02 0.11 0.18 A1203 28.30 25.96 Fe203 1.02 1.76 FeO 3.88 6.55 MgO 2.72 2.50 E20 9.06 10.18 Na20 1.87 1.70 H20 4.07 3.42 98.72 99.77 Mica I is reported to be uniaxial, Mica II, biaxial. 80

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN A sample of one of these phengites was obtained from the Museum of Natural History of Vienna (No. G3399). From the appearance of the material, it seems to be similar to Mica I. However, optical examination of the sample reveals the presence of two phases: a uniaxial type and a normal biaxial muscovite, which are otherwise markedly similar. Weissenberg photographs of the uniaxial variety again show it to possess the 3-layer hexagonal structure, whereas the mica with the normal 2V has crystallized as the common 2-layer monoclinic muscovite type. In view of this admixture of the two varieties in material supposed to be homogeneous, it is doubtful that the chemical differences reported by Wulfing (1886) are significant. Further proof of the composite nature of the sample is given by a powder X-ray photograph which shows lines characteristic of both polymorphs in approximately equal intensities. Scheerer (1862, p. 63) also described two uniaxial micas from a gneiss near Freiberg, Germany, which are high in silica, SiO2 = 50.77 percent and 51.80 percent. Buryanova (1940) reports a yellowish-green mica from a granitic pegmatite of the Korosten pluton in Volhynia associated with albitized orthoclase and lamellar albite that has 2V = 8~ (' = 1.629 + 0.0045). It is pleochroic in shades of light green and occurs with fluorite along cracks in quartz. In the spaces between the sheets of the nearly uniaxial mica is found a lighter green, markedly biaxial mica with strong dispersion of the optic axes (r<v). Buryanova (1940) believes these micas are probably closely related to the potash micas, "gilbert ie and cookeite," formed by the hydrolysis of orthoclase. He further notes that light-green micas found with fluorite have marked dispersion of the optic axes, either r)v or r<v, whereas in similar micas unaccompanied by fluorite the dispersion of the optic axes is not to be observed. From this he concludes that fluorine plays an active role in changing the optic properties of green micas. This hypothesis certainly requires further substantiation before it is acceptable. Clark and Hunt (1915) concluded that the mice in a dolomitic marble from Cockeysville, Maryland, is a muscovite, even though it has the optical properties of a phlogopite. Their conclusion was based on a partial rock analysis which showed a discrepancy in the A1203:SiO2 ratio for phlogopite. A specimen of this mica has been X-rayed and the structure was determined to be that of the 1-layer monoclinic polymorph, with intensities identical with those of other 1-layer phlogopites. Thus it is a phlogopite and not a muscovite with a small 2V. Shibata (1952At p. 135) notes, with respect to a mica from the Bunsen pegmatite, Korea: "White mica associated with pink lepidolite is.___________________________ 81

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN,uinaxial nega-tive ume te m icroScope and i co nsidered as phengite A " An analysis of this materal. by Sh1ibata (19253, p. 150) has SiO2 = 43.24%, MgO 0.90%, and FO + O e2Q3 = 1.93%. It appears therefore that he calls this mica a phengite on tie basis of its uniaxial character and not because of its chermistry. Shibata (1952B) also describes a "lithium phengite" from the LUruchin Mine, Korea and states that it (p. 1553): ".. is optically uniaxial and has high contents of iron and magnlesium. These characteristics coincide with those of phengite and it is to be called for (sic) Liphengite' by its high conftent of lithiimo" Actually this mica also has an essentially normal lithium-bearing muscovite composition, with SiO2 = 47.44%, Mg 1.41% and FeO0 + Fe203 - 1.87% and Li2O -= 02%. These micas need X-ray study. Postel and Adel-ieam (1944) have described a white muscovite of late hydrotlhermal origin f ron the Wissahickon complex in Pennsylvania in which the 2V variecs from 22 to 50~o The explanation they advanced is that random shifts in the structural plane o f the mica may have some bearing on the low and variable 2V. We have been unatle to obtain any of this material for X-ray study but suggest that the sm-all 2V may be due to a rotation of the individual normal muscovoito layers S4itilar to that described by Reusch (1869) and Sugi (1940) 7o. D is cuss 0ion. Clearliy th'ere is a definite relationship between the formation of t2he 3-layer rh -xagon7 polymorph and the high silica content of these muacoviteso All -layrc hexagonal8 muscovites examined by us are of -the high->siiica type (phngites6), bt not nearly all phengites have this structure. Apparently a larger wnumber has crystallized as the 2-layer monoclinic, normal -msc ovite poiy2norpho The only knoJn possible exception to th.is gnearalization is the 5 -layer, lewsilica muscovite of Axelrod and Grimaldi (1949) which has a normal riscovite composition. Its structure, as has been discussed abovTe miay he; interpreted as monoclinic or hexagonal (p.47) There appear to be three possible causes of the polymorphic variat iom: ( 1 ) tv uiing, (2) isomorpkous c(ompositional changes, and (53) envirom.a tal factors As has been noted previouasy, no possible tw-in combinations of the 2-layer monocJlinic Etractlu&e can be arranged to yield the 5-layer hexagonal structure o systemati difft e ie-zces in copos ition are know-a to distinguish the hexagonal frcam the monoclinic phengites o The statement in the Textbook of Mineralogy (Ford, 17952 po 660) regarding miscovite'"2V variable, usually 82

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN about 40~, but diminishing in kinds (phengite) relatively high in silica" is incorrect. Kunitz (1924, p. 383) has stated: "Bei einem weiteren Vergleich dieser Analyse... lassen sich im allgemeinen keine grosseren Abweichungen in der chemischen Zusammensetzung festsletten, die eine befriedigende ErklaIt tt.It rung fur die sprunghafte Anderung im Achsenwinkel geben konnten." He suggests that the optically anomalous nature of the phengites might be related to details in the arrangement and bonding of certain atoms. There is, of course, no certainty as to whether the few available analyses were carried out on completely uniaxial or entirely biaxial material. This also apparently holds for those by Wulfing (1886). However, from the small-scale nature of the optical variation, probably material of both types was involved in each analysis. The phengite of Pagliani (1937) is a good example, for, although the literature indicates that the analysis was performed on material-with a 2V of about 11~, type specimens sent by Professor Pagliani also have a pronounced biaxial phase. At first it was believed that the uniaxial portions of the high-silica muscovites might represent a mica approaching the octophyllite group in composition. This was ruled out however, for the calculated atomic ratios indicate that all of these micas fall definitely within the hep taphyllite (muscovite) group. It does not seem likely, moreover, that the control of the polymorphism can be described entirely to compositional variations, as is the case in the muscovite-lepidolite series (Levinson, 1953). Phengites of appar ently the same composition may or may not crystallize as the 3-layer hexagonal polymorph. Tf the high silica content exercises any control on the structure, it is a permissive rather than a requisite control. The environmental factors are even more difficult to evaluate. Table VII lists the recorded uniaxial or nearly uniaxial phengites together with their modes of occurrence and available information regarding their origin. Although the data are incomplete, it is striking that none of these micas has formed by crystallization from a granitic or pegmatitic magma. They are either of hydrothermal origin or occur in metamorphic rocks, in which their exact origin is unspecified. Although muscovites of secondary origin are characterized, in general, by a somewhat higher silica content (the socalled high-silica sericites) than magmatic muscovites, there are some pegmatitic phengites. These have axial angles of normal size, and those checked by us possess the normal 2-layer monoclinic muscovite structure (e.g., No. 18, Jakob 1925B and No. 54, Jakob 1929C). These relations suggest that if the polymorphism is controlled in any way by, or related in any manner to, the high silica content, an unknown physical factor, furnished by a hydrothermal or metamorphic environment, also must be active before the high-silica muscovite can crystallize as the 5-layer hexagonal polymorph. Any such postulated environmental control must be exceedinigly delicate in order to account for the small-scale intergrowths of the two polymorphs in the same sheet. Foster (1952) has observed a correlation in montmorillonites between degrees of swelling and octahedral substitution. In particular the ______________________85 ___________

TABLE VII: OCCURRENCES OF UNIAXIAL OR NEARLY UNIAXIAL PHENGITES.'AND MUSCOVITES Name Locality Optic Angle Reference Occurrence and origin Alurgite St. Marcel, Italy 2E=0-570 Penfield, 1893 Metamorphic manganese deposit Alurgite East Tyrol, Austria Uniaxial and Meixner, 1939 Metamorphic manganese deposit Biaxial Mariposite Mother Lode, Calif. 2E=0-56~ Knopf, 1929 With ankerite in hydrothermally altered serpentinite lenses, near gold-quartz veins Mariposite Hoggar, Algeria 2V=10~ Sandrea, 1950 Micaceous schist Phengite Fontane, Italy 2V=l8=35~ Pagliani, 197 Mica schist overlying talc lens Muscovite Val Mesocco, Switz. 2V=0~ Jakob, 1929B Alpine cleft vein in eclogite Phengite Mto Mucrone, Switz. 2V=0~ Kunitz, 1924 Gneiss Phengite Adula Mtns., Switz. 2V=0~ Wulfing, 1886 Gneiss Phengite Freiberg, Germany 2V=0~ Scheerer, 1862 Gneiss Yellowish- Volhynia, Russia 2V=8~ Buryanova, 1940 Secondary, in fluorite veingreen mica lets in pegmatitic quartz Muscovite Wissahickon Complexl 2V=22-50~ Postel and Late stage hydrothermal Pennsylvania Adelhelm, 1944 reaction in granite Muscovite Sultan Basin, Wash. 2V=3-15' Axelrod and Veinlets and replacements Grimaldi, 1949 in granodiorite

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN substitution of Fe+35Mg+2 and Fe+2 ions for A1+3 tends to alter the surface energy of the montmorillonite structure. In the case of the high-silica muscovites it seems possible that the distortion, which Hendricks and Jefferson (1959) indicate as the cause for the invariant muscovite structure, may be correspondingly reduced as a consequence of the above-mentioned substitutionS thus permitting other polymorphs to form. It must be emphasized, however, that a decrease in distortion as shown by diminished intensitites of (06) reflections with, odd in the lithian muscovite structure (Levinson, 1953) is not to be observed on ary: she Weissenperg photographs of the biaxial alurgites and phengites. Neverthafless9 it is possible that the polymorphism of the highsilica muscovites may be explained by the fact that these micas, although they approach the ideal muscovite (heptaphyllite) formula, actually are composed of layers with less distortion than those of normal muscovite; the parts that crystallize as the 5-layer polymorph may be relatively undistorted and are thus permitted more choice in stacking arrangement, The cause or causes of polymorphism in muscovite cannot, at this time, be specified. In order to solve this problem, more uniaxial muscovites must be studied and new analyses of material of uniaxial character must be compared with analyses of material that is entirely biaxial. E. Lithian Muscovite In the course of X-ray studies of so-called lepidolites a new variation of the muscovite polymorph has been discovered. A 0-level a-axis Weissenberg photograph of this form is shown in Fig. 6 (compare with Fig. 5). It is apparently confined to "lepidolites" with a low Li20 content and to muscovites with a relatively high Li20 content. For this variation, the term "lithian muscovite" is proposed, to distinguish it from normal muscovitee Lithian muscovite is not to be confused with lithium muscovite, the hypothetical end-member used by Stevens (1938) and Berggren (1941). This variety is not common, having been found in only 10 micas of the approximately 500 muscovites and lepidolites studied. The following characteristics of lithian muscovite illustrate its close structural similarity to normal muscovite. Both have: 1. Space group C2/c. 2. Cell dimensions (approximate; measured on Weissenberg photographs) ao = 5~2; bo = 9.0 A; co = 20.0 A; P = 95~305 3. (06 ) reflections with w odd present. 4. Optic plahe perpendicular to (010).

00.... 02.2....1. /' i Fig. 5. 0-level a-axis Weissenberg photograph of normal muscovite.

REFLECTION / O00~ PLANE /. / i i w:.: / I:1: ":','':': " ~ Fig. 6. O-level a-axis Weissenberg photograph of lithian muscovite. mus covite.

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN The following points are different: 1. Indices are in the normal lepidolite range a = 1.532; = 1.552; = 1.556. 2. Several differences occur in intensity of reflections. The more important intensity differences for (Ok) reflections in normal muscovite and lithian muscovite are given in Table VIII. The observed intensities of normal muscovite are those of Hendricks and Jefferson (1939). The lithian muscovite structure has been found in pegmatitic micas from-: Tordal, Norway; Newry, South Portland and Topsham, Maine, Eight Mile Park, Colorado; Gunnison County, Colorado; Pala, California; Usakos, South West Africa; Kimito, Finland; and Londonderry, Western Australia. TABLE VIII: APPROXIMATE OBSERVED INTENSITIES OF SOME (oki2) REFLECTIONS OF NORMAL MUSCOVITE AND LITHIAN MUSCOVITE Plane Normal Lithian Musc ovite Mus c ovite 020 w a 022 mw vw 026 a vw 045 a vw 061 w vw 065 vw vvw 066 w vvw 067 Va vvw 069 w vvw The data in Table VIII demonstrate that on the basis of the presence of (064) reflections withX odd. lithian muscovite must be considered as having crystallized with the muscovite type structure, but with lesser distortion than in normal muscovite, for most of the (06e) reflections with R odd recorded are extremely weak. This indicates that lithian muscovite approaches more closely the octophyllite micas in structure and composition for, as Hendricks and Jefferson (19399 p. 738) note: these reflections... are absent for the two layer biotite-like micas and none is observed for any of the micas that give (hOj) intensities of the single layer structure (except muscovite)." A "lepidolite" No. 1 analyzed by Stevens (1938), which contains only 2.70 percent Li20, has crystallized with the 2-layer muscovite structure, according to X-ray analysis by Hendricks and Jefferson (1939). Inasmuch as it 88

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN contains such a relatively high Li20 content, Winchell (1942, p. 116) referred to this specimen as a lithium-bearing muscovite. It seemed possible, therefore, that the lithian muscovite structure shown in Fig. 6 was identical with the "lepidolite" analyzed by Stevens (1938) and that Hendricks and Jefferson (1939) neglected to report the intensity differences. Weissenberg photographs of a sample of Stevens No. 1, confirm that this "lepidolite" has the normal muscovite structure as stated by Hendricks and Jefferson (1939). Therefore, Fig. 6 of lithian muscovite illustrates a new variation of muscovite which must contain at least more than 2.70 percent Li20. Rowledge (1945) has briefly described a dozen micas and made partial analyses, including Li20. X-ray studies of portions of these analyzed muscovites and "lepidolites" show that all have crystallized as the normal 2-layer muscovite polymorph. Five of the "lepidolites" have Li20 contents between 2.17 and 2.60 percent. Similarly, three partially analyzed "lepidolites? from Western Australia described by Murray and Chapman (1931) were X-rayed and found to have crystallized with the normal muscovite structure. These micas contain 3.18, 3.24 and 3.32 percent Li20. This information, coupled with the results obtained from Stevens No. 1 (2.70 percent Li20), and the micas described by Rowledge (1945), substantiates the conclusion that much more lithium may enter the muscovite structure without sensible distortion than has generally been realized. In a specimen of mica from Newry, Maine, labeled "lepidolite," the lithian muscovite structure grades into that of normal muscovite. The dozen Weissenberg photographs of flakes from this specimen show the critical reflections to vary in intensity between those of normal and those of lithian muscovite. Spectrographic analysis of this material indicated 4.1 percent Li20. A spectrographic analysis of a lithian muscovite from Tordal, Norway shows 4.3 percent Li20. (These quantitative spectrographic analyses are of a preliminary nature. Future wet chemical analyses may indicate they are in need of revision.) Therefore lithian muscovite probably must contain at least 35.53 percent Li20 and can contain possibly as much as 4.3 percent Li2O. In attempting to explain a form of muscovite with such a high lithium content, one must inquire if (1) it is possible for such a large amount of lithium to replace aluminum isomorphously in the muscovite structure, and if (2) the increased number of lithium atoms are sufficient to cause a reflection, such as (020), which is present in the muscovite structure, to disappear in the lithian muscovite structure. Since the ionic radii of these elements are similar, Al = 0.57A, Li = 0.69A, the substitution is possible and the valence difference may be countered by other substitutions, such as Si+4 for Al+3. However, it is not necessary to have isomorphous replacement of aluminum by lithium. Since the muscovite (heptaphyllite) structure has only 2/3 of its octahedral positions filled, with increasing Li content, the vacant positions gradually can be occupied by additional lithium atoms. This implies 89

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN the existence of a series between heptaphyllite muscovite with a distorted structure, toward octophyllite lepidolite with an undistorted structure. The presence of extremely weak ( 06Q) reflections with. odd in lithian muscovite supports this concept. If these reflections are entirely absent, the mica has the undistorted 2-layer octophyllite type structure. It is concluded, therefore, that Li ions, in terms of as much as 3.3 percent Li20, can enter the muscovite structure without causing any determinable structural variation. More lithium, probably occupying normally vacant octahedral positions, shifts the structure toward that of the octophyllite micas. In lithian muscovite these chanes are already probably of sufficient magnitude to cause a reflection such as (020) to disappear. Actually this hypothesis can only be proved by means of structure factor calculations.which have not, however, been undertaken in this study. The formula of lithian muscovite approximates K2(Al,Li)ca 5.0 (Si6-7, A12-l1) 020 (OHF)4; optical properties of representative specimens are presented on pagel55. The chemical and structural relationship of lithian muscovite to normal muscovite and the various lepidolites is discussed under the muscovite-lepidolite series (p. 139). 90

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN III. OCCURRENCE A. Granites and. Related Rocks Muscovite is a relatively rare constituent of granites, aplites, and a number of other closely allied acid igneous rocks. Its most characteristic occurrence in granites is with biotite in muscovite-biotite granites (also known as two-mica granites and formerly called binary granites). Muscovite granites that contain essentially no biotite also have been reported, but such rocks appear to be unusual and very rare. In addition muscovite occurs in a group of quartz-rich rocks that have been described under various names, including such types as esmaraldite, beresite, alaskite, greisen, and silexite. Most of these rock varieties, however, are either metasomatically altered granitic rocks or represent quartz-muscovite rocks of hydrothermal origin, either replacement or vein material. In aplites, muscovite is reportedly the most widespread mica, being more common than biotite, according to Johannsen (1932). However, it is rarely abundant in aplites. and there exist numerous aplites in which it is entirely absent. Likewise there are others that contain both muscovite and biotite. In muscovite-biotite granites the muscovite usually occurs in close association with biotite, not uncommonly in parallel intergrowth or overgroth with the dark mica. The contacts between the two micas may be those of the crystallographic outline of the biotite, or, more commonly, irregular. These overgrowths or combined overgrowths and replacements have large-scale counterparts in pegmatites. Muscovite also may occur as subhedral to anhedral flakes that cut across biotite flakes and are apparently formed by replacement of the latter. In fact, some iivestigatorq believe that most muscovite in two-mica granites has been formed through the deuteric replacement of biotite. In some cases this is attested to by the presence of by-product magnetite and relict biotite. Granites that contain muscovite are usually microcline granites and do not contain a pyroxene, very rarely an amphibole. B. Pegmatites 1. General. The main rock type and the chief type of mineral deposits in which muscovite occurs is pegmatite of granitic or granodoritic composition. In such deposits mica usually occurs in large crystals, which, if relatively thick and well formed, are known as books. However, many pegmatites of these types may also contain considerable fine-grained muscovite in flakes and scales and also much dense or massive cryptocrystalline muscovite, usually of secondary origin and commonly referred to as sericite. 91

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN On the basis of their internal structure, pegmatites may be divided into three main types: (l)unzonea pegmatites in which there is no differentiation into units of contrasting petrology, (2) zoned pegmatites which are characterized by units of contrasting mineralogy or texture or both which are arranged concordantly, generally parallel with the overall shape and structure of the deposit. They commonly show an approach toward bilateral symmetry with respect to a central unit or core. For a complete discussion of the characteristics of zones and zoned pegmatites, reference should be made to Heinrich (1948) and to Cameronret al. (1949). (3) Complex pegmatites which contain both primary zones and superimposed younger and not uncommonly transgressive or discordant secondary units, such as veins and replacement bodies. Muscovite occurs in all three structural types of pegmatites and occurs in all of the various zone types and secondary units. 2. Deposits. However, deposits of muscovite in pegmatites are restricted and may be classified as follows: a. Disseminated books in unzoned pegmatites: These are uncommonly of major commercial significance and normally, the size and number of books are small. The mica books are indiscriminately scattered throughout the entire thickness of pegmatite. The pegmatites are usually coarse grained masses of quartz, sodic plagioclase and/or microcline and of varying, but usually low, percentages of bookmica. In a few cases, thin border zones may be present. The mica in some may be somewhat concentrated in a very irregular fashion, closer to the walls or along ill-defined quartzose parts of the pegmatite. Pegmatites with disseminated muscovite are usually relatively thin tabular bodies or pods of short extent. b. Wall.zone deposits~ This type probably is the most common, widespread, and valuable from the standpoint of the total quantity of high-quality mica produced from it. In this type, large books of muscovite occur concentrated in the wall zone of pegmatites together with sodic plagioclase, usually oligoclase, and quartz. Minor amounts of microcline and other accessory constituents, such as biotite, also may be present. Wall-zone muscovite deposits may occur in both the hanging-wall and footwall parts of this zone or may be confined to only one side. In some pegmatites, particularly those that dip flatly, there are marked differences in color, structural defects, and concentration between the hanging-wall and footwall muscovites. If color differences are present, the hanging-wall muscovite commonly is ruby, whereas the footwall- mica is green. The hanging-wall concentration may be richer or leaner than that of the footwall, but often is distinctly richer. Not uncommonly within the zone there may be second-order concentrations or shoots, which have been localized along a minor structural feature 92 -

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN of the pegmatite such as a pipe-like roll in the contact. In some pegmatites this shoot has been formed along the trough-like keel at the bottom of the pegmatite where the hanging-wall and -footwall units merge; in others, the wall-zone concentration lies along a sharp upper ridge or crest. Where large wall-rock xenoliths have been enclosed, wall zones of mica also may have been formed in somewhat more interior parts of pegmatites. c. Intermediate-zone deposits: One type of intermediate-zone deposit is the core-margin type; in it, the mica books lie directly adjacent to the central unit, particularly if the core consists of quartz. It is not always certain merely from the position or the shape of this type of deposit whether it is a primary magmatic intermediate zone or whether it represents a replacement unit localized along the core-intermediate zone contact. In flatlying deposits this designation may become particularly troublesome, for in such, the core-margin- unit typically is confined to the -footwall- side of cores, whereas in steeply dipping or vertical pegmatites it is equally well developed on both sides. Not uncommonly the mica books project into the core quartz and, after removal, leave impressions (mica "footprints"). Where mica zones are developed on both sides of a core, there also may be marked variations in quality and concentration. If the core is developed as isolated pods the core-margin zones are also markedly discontinuous. With this type there may also perhaps be grouped the so-called pod deposits, which are characterized by very poorly developed or embryonic zoning and in which the muscovite is more or less restricted to margins of irregularly defined cores, of coarse-textured quartz, or quartz-microcline bodies scattered irregularly throughout finer-grained pegmatite. Another variant is found in the Alabama mica pegmatite district in which disc-like masses of quartz, usually a foot or less long, are scattered rather uniformly in subparallel position throughout finer-grained pegmatite and are fringed by clusters of small mica books. Both the pod and quartz-plate types of deposits represent types transitional between deposits in unzoned pegmatites and those in some zoned pegmatites. d. Central-unit deposits. These are relatively uncommon and show several variants. In some pegmatites of the southeastern states the core consists of "burr-rock," a quartz-rich material with scattered, small,partly aligned muscovite books. In a few pegmatites, for example, several in the Alabama district, a narrow central portion is occupied by a tabular muscoviterich mass which, in some cases at least, may represent the coalescence of the two arms of a core-margin type of muscovite deposit beyond the end of the quartz core. In certain pegmatites, usually the narrower ones in a district, muscovite books are disseminated throughout a central zone of pegmatites that 93

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN are bizonal. In such cases, however, the central unit is not a typical core, but consists of material of wall-zone composition, and thus it is problematic whether deposits of this type should best be referred to as the wall-zone type or the central-zone type of deposit. Where such pegmatites develop quartz cores at lower horizons, the mica not uncommonly remains in the wall zone. e. Fracture-controlled veins and replacement deposits: Thin tabular lenses or vein-like masses as well as larger, more irregular bodies rich in muscovite can be found cutting across the zonal structure of pegmatites in several districts. Some of these secondary units contain considerable quartz, microcline, and plagioclase, as well as book muscovite, whereas others con;i.i essentially of muscovite. In the narrower veins the mica flakes are commtonl, oriented parallel to the plane of the fracture, but in some of the larger, more irregular secondary units the muscovite occurs typically in comb structure4 with the long axes of the blades more or less at right angles to the plane of the unit. Also in these deposits large irregular masses of small, randomll|;. oriented books ("bull" mica) are common, as well as coarse rosettes and irregular clusters of radially or semiradially arranged blades and books. Wedge and A structure are particularly abundant. Fracture-guided or contact-guided replacement masses form large and conspicuous units in pegmatites of some areas, transecting and corroding one or more zones and replacing other zones entirely or in part. Such units commonly consist of albite (including cleavelandite) and muscovite, as well as a large number of various accessory minerals. The secondary mineralization is rarely haphazard, however, and usually is related, in position and intensity to the overall shape and structure of the pegmatite or to position and shape of some of the zones. Closely spaced parallel or subparallel fracture sets, commonly curved, may form the basis for the beginning of the replacement. In other cases the contact between the core and the adjacent intermediate zone has served to localize the initial replacement. If the pegmatite is horizontal or flat-lying, core-margin replacement may be unilateral, confined to the footwall side of the core, exclusively. The intensity of the replacement is commonly strongest along keels or basal troughs of markedly plunging pegmatite bodies, although rolls in hanging-wall contacts also serve, less commonly, as loci of concentration. f. Combined deposits: Some pegmatites contain mica concentrations in two or more types of units; for example, it may occur in the wall zone and in a core-margin unit, or in a pod-like deposit together with fracture-con — trolled muscovite veins. Usually in such deposits there are significant structural and chemical differences between the micas from the two different types of units. I —-----------------— 94 —-------— I

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 3. Exomorphic Mica. In a few districts, relatively large mica books formed by replacement and/or recrystallization occur in the wall rock close to the margin of the pegmatite. In some cases only scattered books are present, whereas in other deposits small local concentrations have been mined. Other occurrences: Graphic to dendritic intergrowths of fine-grained flaky muscovite with various minerals are common in the outer parts of pegmatites in various regions. The most commonly intergrown species is quartz in which the muscovite forms plumose to dendritic aggregates,. Heinrich (1948) has described the occurrence of these widespread and characteristic intergrowths in some Colorado pegmatites. Other less common intergrowths occur with microcline, garnet, and tourmaline. Fine-grained secondary muscovite (sericite) is also abundant, commonly as a replacement of microcline or plagioclase, but also forming at the expense of tourmaline, beryl, topaz, spodumene, amblygonite, garnet, and many other less common species. In some cases this pseudomorphic muscovite may be in the form of relatively coarse scales and may be the variety rose muscovite (Heinrich and Levinson, 1953). C. Hydrothermal Deposits Muscovite is a not uncommon constituent of quartz veins in many districts, particularly in those in which pegmatites are common. Some of these mica-bearing veins are difficult to classify, for they may be somewhat feldspathic and resemble quartz-rich pegmatites. In some examples the muscovite occurs as parallel, coarse flakes near the vein margins and may be at least partly of endomorphic origin, Muscovite occurs in some alpine-type veins believed by Swiss mineralogists to have been formed by precipitation of material leached from wall rocks. Lithium-bearing muscovites (lithionite) are reported as common constituents of tin veins and greisen. Coarse muscovite is generally believed to be indicative of high-temperature environment for the veins or lodes. On the other hand sericitization of wall rock is most strongly developed accompanying mesothermal deposits. In this connection the work of Roy (1949) may have significance. D. Metamorphic Rocks Muscovite is a widespread product of the low-grade regional metamorphism of argillaceous and arenaceous sediments in such rocks as slates, sericite phyllites, chlorite-sericite schists, muscovite quartzites, schistose grits, muscovite schists, and muscovite-biotite schists. Muscovite also 95

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN persists not uncommonly into higher-grade rocks such as sillimanite gneisses and garnet-biotite gneisses. Low-grade regional metamorphism of intermediate igneous rocks produces calc schists with albite, sericite, chlorite and epidote. Similar metamorphism of more felsic types results in albite-sericite schists or muscovite schists. Muscovite-glaucophane schists also are known. Muscovite also forms during the early stages of the contact metamorphism of argillaceous sediments, but with advancing metamorphism, gives way to orthoclase. During retrograde metamorphism, aluminous silicates are commonly transformed into sericite; these include andalusite, kyanite, cordierite, sillimanite, potash feldspar, plagioclase and staurolite ("shimmer aggregace':, Some investigators insist that these changes require the introduction of potassium, whereas others believe that sufficient potassium is available through the decomposition of the potash feldspar. Muscovite is also formed by means of low- to medium-grade metasomatism, notably in contact aureoles close to the granite, where it may replace earlier-formed higher-grade aluminum silicates. Exomorphic muscovite also occurs commonly in wall rocks along the margins of pegmatites, for example, in the Petaca district, New Mexico, in the Black Hills, South Dakota, in the Latah County district, Idaho, and in central Colorado. At the Clein deposit in the Alabama district a sheath of fine-grained muscovite flakes has been developed around a body of aplite intruded into biotite gneiss (Heinrich and Olson, 1953). Muscovite has been synthesized by Noll (1932) at about 2000C. and Roy (1949) has resynthesized decomposed muscovite at 650~C. under high watervapor pressure, showing that the mineral is stable in such an environment to at least this temperature. E. Sedimentary Rocks The micas that occur in sedimentary rocks as authigenic constituents belong chiefly to the illite group and are characterized by a deficiency in K ions, whose positions are partly occupied by hydronium ions (Brown and Norrish, 1952). This investigation has not been concerned with the illite group of minerals. Of course, detrital muscovite and biotite also occur in various clastic sedimentary rocks. F. Variation of Composition with Occurrence Several attempts have been made to plot any systematic relationship that may exist between variation in composition and variation in occurrence. 96

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN The first unsuccessful plot, made by the senior author several years ago, involved a recasting of several hundred muscovite analyses into three components: A = A1203, B = FeO + MgO + MnO + Li20, C = Fe203 + Ti02 + Mn20O. These were recalculated to the basis of 100 percent and plotted on triangular graphs for the following groups: granitic muscovites, pegmatitic muscovites, hydrothermal muscovites, metamorphic muscovites and sedimentary muscovites. Another similar attempt used as the three components: A = FeO + MgO + MnO, B =A10 + Li20, C = Si02 + Fe203 + TiO2. This resulted in a better distribution of the points, but, in general, the spread within each paragenetic group approached the variational range for the entire collection of analyses. Some trends, however, were apparent from the plot: (1) Pegmatitic muscovites show the greatest range of compositional variation. (2) Most of the high-silica muscovites (phengites) are formed under conditions of regional metamorphism. (3) Fine-grained muscovites ("sericites") of lowtemperature origin tend to have potassium slightly deficient and water slightly in excess. (4) F and Li appear almost entirely in pegmatitic and in hightemperature hydrothermal muscovites. 97

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN IV. PEGMATITIC MUSCOVITE A. Structural Defects Many muscovite books are imperfect crystals, containing a variety of macroscopic imperfections. 1. Reeves: Shallow, closely spaced, parallel grooves or narrow corrugations that interrupt the cleavage continuity to various depths. The reeves are oriented parallel with (110) and (010). Several combinations of lines are common: A structure —2 sets of reeves in a V-pattern, intersecting at 60~, double A structure —3 sets of reeves, joining at a common center at angles of 60~; herringbone structure ("fishback" or "fishbone") —2 wide sets of reeves at 1200, bisected by a third, prominent, narrow set; radial reeves —at 60~, six sets radiating from a common center. 2. Ruling: Sets of parallel parting planes, cutting the cleavage at 67~ and parallel with (110) and (010). One, two, or three sets may be present, resulting in strips, rhombs, or diamondshaped pieces. 3. Fractures: Usually curved, sharp breaks that are not crystallographically oriented. May be semi-radial from large inclusions, particularly quartz. Extremely fine cracks that appear only as the book is split are called hair cracks. 4. Wedge structure: The books are wedge shaped in cross section, that is, the sheets or laminae are tapering. Wedge and A structure not uncommonly occur together. 5. Warping: Ridges or waves of varying dimensions and extent, from minute, abundant crenulations to broad warps or bucklings. Cleavage steps are sharp minute flexures that result in a stairstep surface on cleavage fades. 6. Block structure: Some mica books consist of slightly disoriented, discrete units whose orientations deviate a few degrees from those of their neighboring blocks. The surfaces of the blocks extend inward to the crystal center. This is one of the causes of imperfect cleavage across the entire book ("locked" or "tangled" mica). Concentric structure is a 98

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN variation of this in which small mica plates are so misoriented with respect to their neighbors that the aggregate forms a curving, hemispherical mass. One way in which such a structure forms is by deposition over the curved surface of a sheave of cleavelandite blades. 1. Muscovlte-Biotite Overgrowths and Intergrowths. Muscovite and biotite commonly occur together in composite books or strips, in which the two micas have their cleavage planes parallel or nearly so. In some types one species surrounds or nearly encloses the other, and the boundary between the two may be a euhedral crystal outline or it may be highly irregular. Commonly, biotite is enclosed in muscovite. In a few of these the enclosed crystal is markedly pyramidal in habit and tapers parallel to c, so that at one end of the crystal the central mica occupies a much smaller part of the sheet than at the other. Normally the included plate or crystal is bounded by prism and pinacoid faces. A few rare examples consist of a euhedral muscovite core, an intermediate biotite strip, and a muscovite margin. In other composite books there is no regular pattern formed by the two micas. The cleavages of the two are still generally parallel but the distribution of the two types either across the cleavages or parallel to c is not systematic. Some examples of these overgrowths and intergrowths are described below. The only nonpegmatitic occurrence of muscovite and biotite intergrowths was reported by Johannsen (1938). A nepheline-bearing diorite called dungannonite contains small amounts of yellow to greenish-brown biotite intergrown with muscovite. However, subparallel intergrowths of the two micas are not uncommon in two-mica granites. Composite books of pegmatitic biotite and muscovite, as much as 12 inches across, have been reported from the following representative localities: Southeastern Piedmont District: The Mitchell Creek Mine, Upson County, Georgia, has produced muscovite-biotite intergrowths of several varieties. Furcron and Teague (1943) report sheet muscovite books wrapped by sheet biotite, biotite books included in muscovite, and muscovite included in biotite. Both species have a common cleavage plane. Lester (1946) found two types of intergrowths at this locality. The first is a tight intergrowth in which the two micas possess a common -cleavage plane. Thin sheets of this type can be cleaved as easily across the line of contact as within the individual micas. The second type is a loose intergrowth with straight boundaries; the two micas possess common planes of cleavage, but tend to separate readily when cleaved across the contact. Rum-colored muscovite showing heavy A structure and containing six-sided inclusions of biotite has been reported by Furcron and Teague (19453) 99

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN from the Peters Mine in Monroe County, Georgia. They also found a minor occurrence of biotite crystals included in sheets of muscovite from the following localities in Georgia: Dick Fletcher Mine, Monroe County W. M. Gooch Mine, Lumpkin County W. A. Sullivan Mine, Union County Chapman Mine, Elbert County. In Mitchell County, North Carolina, Sterrett (1923) found crystals and sheets of biotite with included crystals of muscovite and vice versa. The two micas generally occur in parallel intergrowth and have a common cleavage. Studies of a specimen of biotite enclosing a rhombic crystal of muscovite showed the two micas to have approximately parallel percussion figures and optic planes. Some pegmatites of the Spruce Pine District, North Carolina, contain biotite intimately mixed with oligoclase and muscovite. Maurice (1940) reports that, in the outer zone of the pegmatites, muscovite frequently forms an outer zone around biotite crystals. Thin, flaky sheets or laths of biotite occur as veinlets in parallel intergrowth with muscovite. The laths are usually less than 4 inches long and 1/16 inch thick. Sterrett (1925) also reported parallel intergrowths of biotite and muscovite from the Hamilton Mine, Ashe County, North Carolina, the Big Ridge Mine, Haywood County, North Carolina, and the Chalk Hill Mine, Macon County, North Carolina. Pogue (1911) found six-sided plates of biotite enclosed in muscovite at the Buck Creek deposit, Macon County, North Carolina. Examples also are cited by Jahns- and Lancaster (1950). Northeastern Uhited Statesr Several occurrences of biotite-muscovite intergrowths have been reported from New England pegmatites. However, none of the descriptions are detailed; apparently the occurrences are not common. Generally, the micas intergrow in parallel position with a common basal cleavage. The reported localities are: Island Mine, Creshire County, New Hampshire Sterrett (1923) and Olson (1942) Holden Mine, Granton County, New Hampshire Sterrett (1923) Patten Mine, Granton County, New Hampshire Sterrett (1923) Old Lithia Mine, Chatham,Connecticut Shannon (1920) 100 _

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Middletown, Connecticut Pogue (1911) Philadelphia, Pennsylvania Hall (1885) Lenni, Delaware County, Pennsylvania Hall (1885) Western United States: Biotite-muscovite intergrowths from the Western states have been described from a few pegmatites. Hanley et al. (1950) describe green-gray, heavily stained muscovite from the Rosemont Mine, Micanite District, Park and Fremont Counties, Colorado. The muscovite is ruled, ribboned, has pronounced A structure, and is commonly intergrown with biotite. The same authors also report blades of intergrown muscovite and biotite as long as 6 feet from the School Section Mine, Eight Mile Park, Fremont County, Colorado' Other scattered occurrences were at the following localities: Levi Anderson Mine, Latah County, Idaho Sterrett (1923) Custer County, South Dakota Pogue (1911) Canada: Spence (1930) reports parallel intergrowths of muscovite and biotite from dikes of the lower St. Lawrence area. The following three varieties of intergrowths have been found at the Pied des Monts mica mine near Murray Bay, Quebec: (1) core of muscovite surrounded by a rim of biotite, (2) books that are part muscovite and part biotite, the boundary running diagonally across the cleavage sheet, and (3) alternate layers of muscovite and biotite. Minas Gerais, Brazil: Occurrences of intergrown biotite and muscovite were noted by Pesora et al. (1950). Generally, the intergrowths are in parallel position and occur in the border zones of the pegmatites. Germany: Linck (1901) reported regular and irregular intergrowths of biotite and muscovite from Veltlin. He found that the rays of a percussion figure near the boundary would cross the boundary without interference. The two species of micas have a common cleavage plane, but the planes of the optic axes are at an angle of 60~. Scharizer (1886) described regular overgrowths of muscovite and lepidomelane from the pegmatites of Schuttenhofen, Germany. A six-sided 101

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN crystal of lepidomelane is completely surrounded by a rhombic crystal of muscovite. Both species have a common basal cleavage. The axial plane of the lepidomelane is parallel with the long ray of a percussion figure, whereas the axial plane of the muscovite is normal to the long ray of the percussion figure; the two optic planes are therefore it an angle of 600. Kimito, Finland. Pehrman (1945) reports weathered books of interlayered biotite and muscovite from the granitic pegmatite of Kimito. Orientation Observations: A group of specimens of intergrown muscovite and biotite was selected from the Michigan mica collection for optical study. An attempt was made to find the relationship between percussion figures and optic planes in the two species. In all cases, the percussion figures are essentially parallel. The greatest deviation from parallelism was found to be 9~. The most common angle between the optic planes of the two species is approximately 60~. However, angles of 0~ and 90~ were also noted. Generally, the line of contact between species is sharp. In a few specimens, the contact is ragged or feathered. In no specimens was a gradation found between muscovite and biotite. The data obtained are listed below. Specimen Angle Between Main Ray Number Locality of percussion Figures Optic Planes Perpendicular 85 Putnam, Haywood County, 0~ North Carolina Optic Planes Parallel 149 Mauldin Road prospect, Upson 5~ County, Georgia 1393 Location unknown 0 Optic Planes at an Angle of 60~ 47 Ledford Cove, Macon County) 9~ North Carolina 48 Ledford Cove, Macon County} 0~ North Carolina 68 Big Ridge, Haywood County) 6~ North Carolina 102

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Specimen Angle Between Main Ray Number Locality of Percussion Figures Optic Planes at an Angle of 60~ (cont.) 84 Lower East Fork, Haywood 4~ County, North Carolina 153 Mitchell Creek, Georgia 3~ 177 M and G, Alabama 6~ 282 Eight MilePark, Fremont 0~ County, Colorado 1391 Hebron, Maine 0~ 1392 Delaware County, Pennsylvania 0~ 1394 North Carolina 00 Jahns and Lancaster (1950) state (p. 15) that the pressure and percussion figures of the inclusions commonly are oriented normal to those of the host. Conclusion. Grigoriev (1936), who studied the interrelations of muscovite and biotite in some detail, concluded that the similarity in crystal structure, cleavage, and crystal form allows for the possibility of (1) regular overgrowths of one mica by another, (2) oriented intergrowths:, and (3) possible mutual replacement. The relations of biotite to muscovite in the composite crystals can be grouped as follows: 1. Euhedral biotite plates enclosed in muscovite (common). 2. Euhedral to subhedral pyramidal biotite enclosed in muscovite (uncommon). 3. Euhedral muscovite plates enclosed in biotite (rare). 4. Euhedral muscovite in euhedral to subhedral biotite enclosed in muscovite (very rare). 5. Anhedral strips and plates of biotite enclosed by muscovite or irregularly interErown with muscovite (common). 6. Alternating sheets of muscovite and biotite (rare). 103

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN B. Inclusions 1. Magnetite and Hematite. Magnetite and hematite are by far the most common and abundant included minerals. They occur as very thin (usually <0.01 to 0.1 mm) plates, needles, skeletal crystals, and lattice-work aggregates (Frondel and Ashby, 1937) oriented parallel to either percussion or pressure arms. Concentrations of the iron oxides occur usually in the central parts of books, whereas margins are usually clear. Most of the hematite inclusions appear to have formed as the result of exsolution. At the initial temperatures of crystallization the muscovite structure is capable of retaining considerable iron isomorphously. With a decrease in temperature the tolerance is decreased and the iron is ejected and segregated as hematite units The outer parts of crystals, which formed at lower temperatures than the interior, could hold less iron originally and thus were able to retain all of this lesser amount of iron upon cooling. In books with such inclusion zoning, the muscovite of the clear margins not uncommonly has higher refractive indices (more Fe2 and/or Fe3) than that of the core from which the maximum amount of iron has been removed through exsolution. In pegmatites where muscovite and biotite occur together in the same zone, iron-oxide inclusions are rare in the muscovite, probably because the normally earlier crystallization of the biotite abstracted most of the available iron from the pegmatitic solution. 2. Other Minerals. Numerous other mineral species occur as thin inclusions in muscovite (Frondel, 1936, 1940; Jahns and Lancaster, 1950). The shape and orientation of many such mineral plates or needles have been controlled by the principal crystallographic directions of their host, and the inclusions are primary. Other inclusions display little or no orientation and are randomly intergrown with the mica either as films along cleavages or even transecting the books and thus breaking the cleavage continuity. Such mineral inclusions may be either primary or secondary. Books marked by abundant inclusions of grains or rods of quartz, albite, apatite and a few other species are termed aBand-pitted or gritty. 3. Other Materials. Gas bubbles of primary origin, or minute secondary air pockets form inclusions in some types of muscovite. Green amorphous patches and films of undetermined organic material are included in weathered muscovite. C. Variations in Individual Pegmatites 1. Concentrations. Despite the fact that muscovite can be a constituent of any pegmatite zone and of replacement units, most concentrations ------------------------ 1 4 —------------

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN occur in: (1) wall zones, (2) intermediate (core-margin) zones, and (3) replacement units. Unzoned pegmatites that contain disseminated books also are of economic significance in some districts. Individual pegmatites may contain concentrations in more than one unit, although if muscovite occurs in more than one unit, normally it will be much more abundant in one than in the other. The concentration of mica in a single-zone is not always uniform. In some pegmatites, the hanging-wall part of the wall zone will alone contain muscovite concentrations. Exceptionally the footwall. counterpart alone may have them. In other pegmatites concentrations within the wall zone are localized along rolls or bulges in the contact, so that mica shoots have been developed. The margins of xenoliths may serve as a false wall along which a concentration of wall-zone mica is repeated. 2. Structural Defects and Color. Within an individual zone the color, degree of staining (iron-oxide inclusions), and structural defects are usually uniform, although numerous exceptions are known. The Gregory Mine in the Franklin-Sylva District of North Carolina, for example, contains ruby muscovite in the hanging-wall part of the wall zone, green muscovite in the footwall. part of the same zone, and where the two merge around the keel of the pegmatite, the books are color-zoned, some with green cores and ruby margins. Commonly, if two different colored micas occur in the same pegmatite, that of the wall zone is brown or ruby, whereas that of the core-margin zone is green. If the wall-zone mica is green, then if any core-margin muscovite is present, it also is green or silvery. Brown or ruby core-margin muscovite does not occur with green wall-zone muscovite. MuscoVite of fracture fillings and replacement units commonly is green, silvery, or almost colorless. Rose muscovites are of late replacement origin. Muscovite-biotite intergrowths are essentially confined to outer zones of pegmatites, and, generally, wall zones containing green mica contain little or no biotite. Moreover, in wall zones that contain both biotite and ruby or brown muscovite, the latter normally is relatively free of iron-oxide inclusions. Inclusions of iron oxide usually are more common and abundant iri green than in ruby muscovite. The muscovite of wall zones more commonly is in relatively small, flat books, free of A or herringbone structure but may contain a relatively large number of small mineral inclusions (except iron oxides) and unoriented fractures. In contrast, core-margin mica is characterized more commonly by relatively large size, widespread A or herringbone structure, and iron-oxide staining. Mica from replacement units or fracture fillings shows much wedgeA or wedge herringbone structure and extreme variation in book size. Intermediate zones or pegratites in which microcline is the dominant feldspar also contain strongly reeved muscovite. 105

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 35 Composition. Information is now available on the compositional homogeneity of single muscovite books, on the variation of books within the same zone of an individual pegmatite, or on the variation between muscovites from different zones within individual pegmatites. Table IX shows two analyses on sheets from opposite ends of a single thick book of muscovite from the Tilley pegmatite in the Franklin-Sylva district, North Carolina. They show remarkable uniformity. In Tables X through XVII are recorded chemical analyses of different books from the same zone of various pegmatites. TABLE IX SPECTROCHEMICAL ANALYSES* OF MUSCOVITE SHEETS FROM OPPOSITE ENDS OF A SINGLE THICK BOOK (Tilley pegmatite, Franklin-Sylva district, N. C.) 436 a 436 b Fe2O3 3.5 3.7 MgO 1.5 1.5 MnO.02.02 Ti02.30.27 CaO.002.004 BaO.11.075 Na20 1.2 1.25 Ga203.015.015 Cr203.0.0005. V205.008.005 Absent: Li20, Rb20, Cs20, SrO, Sc203, SnO2, Co203and F. * Analyses by C. E. Harvey 106

TABLE X SPECTROCHEMICAL ANALYSES* p ID:FFERBENT' COYITES FROM TEE SAME ZONE WITHIN TEE BIG RIDGE EGMATITE (Franklin-Sylva diatrictf K.t C.) 67 70 72 73 75 76 77 81 82 Volk** Fe203 2.2 2.1 2.2 2.2 2.4 2,0 1.8 2.1 2.1 2.09MgO 1.5 1,2 1.2 1.3 1.5 1.3 1i15 1.2 1.3 1.57 MnO.04.04 06.04 4 05.4.0 4 Ti02.16.26.27.13.15.21.21,17.16 0.89 CaO..002 oo009 015 003.02 0oo35 004.01,0 0o80 SrO --.00300.002.00 2..002.003.002.002 > BaO 3.9.52.44.47.44.47.46.43 o41 Na20.64.72.60.60 *55.97 67.75.81 1.59 Li20 - -- -- -. --- - - Rb20 Cs20 -- -- - - _ -- Ga203.017.017.02.02.018.017.02.022.015 Sc203.004.005.006.003.004.005.006.004.003 o0203.000 ooo04.0004.0004.0004.ooo4.0005.ooo005.0004 Cr203.0001.0002.0002.0002.0003.0001.0002.0002.0002 Sn02.0007.0007.001.001.0008 0007.001.002.0008 V205.003.006.006.004.005.005.006.005.003 F - - - -- -- —...... - - m m —,.. _ _ * Analyses by C. E. Harvey * Am. Min",24, 255-266, 1939 f Total iron

TABLE XI SPECTH)CEEMICAL ANALYSES* OF DIFFERENT MUSCOVITES FROM THE SAME ZONE WITHIN THE POLL MILLER PEGMATITE (Franklin - Sylva district, N.C.) 54 55 56 57 59 362 Fe203 2.3 2., 2.0 23, 2.5 2.7 MgO.95 1.2 1.2.90 1.2.85 MnO o13.12.03.10.4.03 TiO2.07.08.21.05.08.07 CaO.002.001.002.002.002.0007 SrO BaO.06.002.17.005.05.003 Na2QO 90.85.78.90.80 1,2 Li20 Rb20 Cs2- -'a"o~.026.022.018.024.015.017 Qc203 -- -.003 -.001 Co203.0002 -.0003 --.0004 Cr20 -- - -.0001 Cr203 SnO2.006.003.0007.005.001 V205.007.0001.002.002.0004 - *jAnalyses by C. E. Harvey 108

TABLE XII SEECTROCHEMICAL ANALYSES* OF DIFFERENT MUSCOVITES FROM THE SAME ZONE WITHIN THE TITT Y PEGMATITE (Franklin-Sylva district, N. C.) 18 20 21 22 23 Fe203 2.3 3.7 4.2 3.4 3.3 MgO.78 1.1 1.3 1.4 1.15 MnO.08.08.024.06.11 TiO2.09.20.13.16.15 CaO.002.002.002.004.43 SrO.003 -- --.002 BaO.23.10.03.08.09 Na20.66.68 1.5 3.65 Li20 Rb20 Cs20 - Ga203.017.018.016.02.022 Sc203.003.002 -- 003.002 Co2O3.0003.0004 --.0005.0004 Cr20O.00.0008 --.0003.0004 SnO2.0007.001.004.002.002 V205 -.01 --.006.01 * Analyses by C. E. Harvey 109

TABLE XIII SPECTROCHEMICAL ANALYSES* OF DIFiRENT MUSCOVITES FROM THE SAME ZONE WITHIN THE RUBY, NORTON, SHEEP MOUNTAIN, AND DOC NICOLS PEMATITES (Franklin-Sylva district, N. C.) Ruby Norton Sheep Mtn. Doc Nicols 60 61 132 133 15 16 2a 2b Fe203 2.6 2.1 4.4 4.2 2.3 2.2 3.0 5.3 MgO.92.97 1.4 1.25.85 75.95 1.0 MnO.017.012.09.06.018.04.04.04 TiO2.15.21.13.20.42.38.07.05 CaO.0007.0005.002.001.001.001.004.002 SrO BaO.04.04.01.055.27.28.008.006 NaO2 1.7 1.2 1.0 1.05.91.80.83.62 L20 -' Rb20 Cs20 Ga203.017.016.014.014.013.015.021.022 Sc203 -- - -- --.003 Co2 - — 2.0006.0006 -.0002.0003 Cr203 --.0002 - --.0004.0004 Sn02.004.002 - - --.0007.003.001 V205.003.004 -.001.024.036.009.002 F - --.20.10 -- -- -- - * Analyses by C. E. Harvey 110

TABLE XIV SPECTROCHEMICAL ANALYSES* OF DIFFERENT MUSCOVITES FROM THE SAME ZONE WITHIN THE MITCHELL CREEK AND JOHNSON PEGMATITES (Thomaston-Barnesville district, Georgia) -__ _ Mirtchell CreeLk _ " ___. _ Johnson 150 151 - 154 146 -147 Fe203 1.7 1.5 1.6 2.8 2.8 MgO 1.55 1.25 1.5 1.05 1.05 MnO.03.029.07.029.053 TiO2.30.29.22.14.15 CaO.002.003.003.0oo6.001 SrO -- -.002- - BaO.14.15.13.004.004 Na20 1.25 1.75.80 1.2 1.25 Li20 Rb2O - - Cs20 Ga203.014.015.02.02.02 Sc203 - -.5 - Co20 -- --.0004 Cr203.001.002.004 SnO2 -- -.014.002.002 V205.011.018.03- - F -- - --.25.25 * Analyses by C. E. Harvey 111

TABLE XV SPECTROCHEMICAL ANALYSES* OF DItFERENT MUSCOVITES FROM THE SAME ZONE WITHIN THE SCHOOL SECTION PEGMATITE (Eight Mile Park, Colorado) 279 281 290 295 300 305 Fe203 3.1 3.3 3.4 3.1 2.9 3.2 MgO.91.90.92.85.75.96 MnO.07.055.05.10.049.14 TiO2.048.12.23.05.11.11 CaO.002.003.002.001.002.005 SrO BaO --.0006.001 --.001.002 Na20.94 1.6 1.4 1.15.74.97 Li2O -- -.19 --.22 Rb20 Cs20 Ga203.016.018.019.018.014.017 Sc2O3.004.005 -.004 Co203 -- - Cr203 Sn02 --.004.005.002 V205 --.001.003 - - F.62.39.25. 73.27.87 * Analyses by C. E. Harvey 112

TABLE XVI SPECTROCHEMICAL ANALYSES* OF DIFFERENT MUSCOVITES FROM THE SAME ZONE WITHIN THE MEYERS QUARRY PECGMATITE (Eight Mile Park, Colorado) 312 316 319 320 324 Fe203 2.5 3.4 2.3 2.5 3.6 MgO.16.90.13 1.1 1.1 MnO.21,06.12.07.07 TiO2.009.07.008.014.08 CaO.005.002.002.003.002 SrO BaO.0007.0008 --.0002.0005 Na20.80.62.75.88.70 Li20.52.06.23.11.13 Rb2 - Cs20 Ga203.024.021.025.025.022 Sc203 --.002 -- --.005 C 020 -.0003 --.0002.0003 Cr203 - -- -- --.001 SnO2 -.001 -- --.002 20 - - - --.006 F.53.18.35.27.17 * Analyses by C. E. Harvey 113

TABLE XVII SPECTROCHEMICAL ANALYSES* OF DIFFERENT MUSCOVITES FROM THE SAME ZONE WITHIN THE GLOBE PEGMATITE (Petaca district, New Mexico) 209 210 Fe203 4.2 3.2 MgO.03.01 MnO.49.81 TiO2.15.045 CaO.001.002 SrO BaO.003.0008 Na20 1.45 1.1 Li20.59 354 Rb20 Cs20 -- - Ga2O3.04.035 Sc2O3 3 Sc203 Co203 Cr203 - SnO0.004.005 V205 F 1.48.98 * Analyses by C. E. Harvey 114

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN These comparisons show that, in the main the variation in composition among different books from the same zone in a single pegmatite is small. This is particularly true of those pegmatites of the eastern districts in which the mica-bearing units are continuous throughout a large part of the pegmatite. In those deposits of some of the western districts, in which the mica-bearing units are markedly discontinuous, the variation of the mica from one isolated part of a unit to that of another is considerably more pronounced. In Tables XVIII and XIX are presented some comparisons of micas from different units or micas of different colors from the same pegmatite. In all cases the differences in color are a reflection of marked differences in chemical composition. 115

TABLE XVIII SPECTROCEEMICAL ANALYSES* OF MUSCOVITES FROM THE GREGORY PEGMATITE (Franklin-Sylva district, N. C.) green Zoned book from keel footwall green ruby zone core margin 4~ 5a 5b Fe203 1.8 2.7 5.0 MgO.71.91.33 MnO.03.03.05 TiO2.03.01.014 CaO.002.001.01 SrO --.001.001 BaO.30.22.002 Na20.85.66.58 Li20 Rb20 Cs20 - Ga203.011.01.016 Sc203 - Co203 Cr203 SnO2 --.003.011 v20o --.002.0008 * Analyses by C. E. Harvey 116

TABLE XIX SPECTROCHEMICAL ANALYSES* OF MUSCOVITES FROM KISER PEGMATITE (Franklin-Sylva district, N. C.) wall zone core margin ruby een 44a 44b 45 Fe203 2.6 2.6 3.5 MgO 1.1 1.4 1.3 MnO.05.07.12 TiO2.20.06.05 CaO.002.0006.001 SrO.003 -- BaO.52.07.37 Na2O.77 -70.75 Li2O Rb2O - Cs20 - -- -- Ga20.017.02.018 Sc203.004.002 Co203.0.0005.0000003 Cr203- -- SnO2.001.001.001 V205.002.0004.0004 F * Analyses by C. E. Harvey 117

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN D. Variations in Pegmatite Districts 1. Occurrence and Color. Within a single district the pegmatites commonly contain mica concentrations in several types of units, but usually one type of deposit predominates. Throughout the southeastern states the bulk of the larger mica books occurs in wall-zone concentrations. In the Alabama district, for example, disseminated and quartz-plate types of deposits occur in over 50 percent of the pegmatites, core-margin in nearly 20 percent. The wall-zone type is found in only about 10 percent, yet over 50 percent of the total production has come from the last. In the Franklin-Sylva district about 30 percent of the pegmatites contain wall-zone deposits, a little more than 20 percent are of the disseminated type, and about 35 percent are of the coremargin type. Commonly in a single district one color type predominates in the mica, though most districts show considerable variation. In most of the Piedmont districts of North Carolina, South Carolina, Georgia, and Alabama the dominant color is ruby. In the Ridgeway-Sandy Ridge district of Virginia and North Carolina green mica is common. In other southeastern districts there appear systematic color variations which can be correlated with the position of the pegmatite with respect to its parent batholith. These occur in the Spruce Pine, Franklin-Sylva, Ridgeway-Sandy Ridge and Hartwell districts (see Heinrich, 1953). In the Latah County, Idaho, district ruby micas predominate. Light brown types are common in Montana. In the Eight Mile Park area of Colorado the core-margin mica is very pale ruby, whereas that of fracture-controlled units is silvery green. Most of the Petaca, New Mexico, mica is green. 2. Composition. Tables XX through XXVIII show the variation in minor and trace elements in muscovites from nine districts. Fe203 averages show considerable variation: high in Spruce Pine, Petaca and Eight Mile Park districts; intermediate in Franklin-Sylva, low in Shelby-Hickory, ThomastonBarnesville, Alabama, and Montana, and very low in Idaho. MgO averages show much less -variation, but are conspicuously low for Petaca micas and somewhat low for Idaho micas. Conversely, the Petaca district micas have high MnO contents. TiO2 averages show less variability and are lowest in micas from Montana. CaO averages are very similar except for that of the Franklin-Sylva micas, which is relatively high. Sr is absent in four districts, and highest in Franklin-Sylva micas. BaO averages vary widely: high in Franklin-Sylva, Spruce Pine, and Alabama, low in Shelby-Hickory, Thomaston-Barnesville, Montana; and very low in Idaho, Petaca, and Colorado muscovites. Na20 averages show less variation; that for the Georgia district is somewhat higher than the norm. Li20 is absent in four districts, uncommon and _ 118 _____ —---

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN low in three (Spruce Pine, Alabama, Idaho), common and moderate in Eight Mile Park and very common and highest in Petaca micas. Rb20 and Cs20 were not detected in muscovites from any district. Ga203 is present in all the analyzed muscovites; the averages are very uniform, except for the Petaca district which is about double those of the others. Sc203 is absent in the Idaho micas, very low in Petaca micas, relatively uniform for other districts. Co203 is absent in one district, very low in most and highest in Spruce Pine muscovites. Cr203, absent in one district, is uncommon and very low in all others. SnO2 has a remarkable concentration in the Idaho district, is relatively high in the Petaca district, and is fairly uniform for the rest. Idaho, Petaca, and Montana districts are characterized by low V205. F is low in Franklin-Sylva, Shelby-Hickory, Georgia, and Montana; absent in Alabama; moderate in Colorado, Spruce Pine, and Idaho; and exceptionally high and invariably present in Petaca muscovites. These variations in averages serve to show that certain elements, namely, Fe,Mn,Ba,Li,Sn, and F, show marked variation from district to district. These variations are ascribable to two causes: (1) fundamental differences in the parent magma from which the pegmatite liquids were withdrawn —variation by provinces; (2) differences in crystallization stage of the muscovites - some are magmatic, others hydrothermal —variation by differentiation. -119 - -

TABLE XX * MINOR AND TRACE ELEMENTS IN MUSCOVITES FROM PEGMATITES (Franklin-Sylva District, North Carolina, and north Georgia) Oxide Present in % of Range Average Samples Fe203 100 1.6 - 5.1 2.77 MgO 100.33 - 1.7 1.12 MnO 100.009 -.17.058 TiO2 100.01 -.52.166 CaO 100.0005 -.43.015 SrO 33.001 -.005.0009 BaO 100.0004 -.62.193 Na20 100.33 - 1.7.826 Li20 - Rb2O 0 --- _- Cs20 0 Ga203 100.010 -.026.017 Sc203 44.001 -.006.0014 Co203 59 o0002 -.0008.0002 Cr203 49.0001 - o002.0002 SnO2 74.0007 -.011.0018 V25 8.0001 - o036.0066 F 8.13 -.42.029 * Based on spectrochemical data by C. Eo Harvey 120

TABLE XXI * MINOR AND TRACE ELEMENTS IN MUSCOVITES FROM PEGMATITES (Spruce Pine District, North Carolina) Oxide ee Range Average Samples Fe203 100 1.6 - 5.1 3.78 MgO 100.25 - 1.9 1.05 MnO 100.015 -.09.066 Ti02 100.01 -.30.11 CaO 100.0005 -.004.0013 SrO 9.003.0002 BaO 100.0005 - 1.10.215 Na20 100.42 - 1.4.917 Li20 9 ~27.02 Rb20 0 - Cs20 0 Ga203 100.013 -.033.021 Sc203 36.001 -.003.0003 Co203 28.003 -.005.0007 Cr203 9.003.0002 Sn02 72.001 -.007.002 V205 82.001 -.12 003 F 54.10 -.73.17 * Based on spectrochemical data by C. E. Harvey 121

TABLE XXII * MINOR AND TRACE ELEMENTS IN MUSCOVITES FROM PEGMATITES (Shelby-Hickory District, North Carolina) Oxide Present in % of Range Average Samples Fe203 100 1.6 - 2.7 2.08 MgO 100 74 -.98.88 MnO 100.013 -.08.035 Ti02 100.069 -.19.119 CaO 100.0005 -.002.0012 SrO 0 BaO 100.005 - 09.003 Na20 100.70 - 1.05.89 Li20 0 Rb20 0 - Cs20 0 Ga203 100.01.032.019 Sc203 20.006.0012 Co203 20.0004.00008 Cr203 0 _ -- Sn02 80.002 -.006.005 V205 40.005.002 F 20.12.024 * Based on spectrochemical data by C. E. Harvey 122

TABLE XXIII * MINOR AND TRACE ELEMENTS IN MUSCOVITES FROM PEGMATITES (Thamaston-Barnesville District, Georgia) Present in % of Oxide Range Average Samples Fe203 100 1.5 - 2.8 2.08 MgO 100.68 - 1.35 1.08 MnO 100.012 -.10.037 TiO2 100.03 -.40.19 CaO 100.0006 -.003.0016 SrO 10.002.0002 BaO 100.003 -.15.058 Na2O 100.80 - 1.75 1.24 Li2O0 0 - Rb20 0.- Cs20 0 —. _ Ga203 100.013 -.025.018 Sc203 20.002 -.005.0007 Co203 20.0004 -.0006.0001 Cr203 40.0008 - o004.0008 SnO2 50.002 -.014.003 V205 50.0007 -.03.0065 F 40.15.2.8 093 * Based on spectrochemical data by C. E. Harvey 123

TABLE XXIV * MINOR AND -TRACE ELEMENTS IN MUSCOVITES FROM PEGMATITES (Alabama District) Oxide Present in of Range Average Samples Fe203 100 1.4 - 2.4 1.88 MgO 100.88 - 1.5 1.13 MnO 100.014 -.10.052 TiO2 100.06 -.21.11 CaO 100 001 -.004.002 SrO 15.002.0003 BaO 100.001 -.68.166 Na20 100.38 - 1.6.84 Li20 29.07 -.26.047 Rb2O 0 --- Cs20 0 Ga203 100.011 -.022.017 Sc203 57.003 -.005.002 C0203 29.0003 -.0004.0001 Cr203 43.001.0001 SnO 71.004 -.012.004 V205 71.001.017.005 F 0 * Based on spectrochemical data by C. E. Harvey 124

TABLE XXV * MINOR AND TRACE EIEMNTS IN MUSCOVITES, FROM PEGMATITES (Latah County District, Idaho) Present in % of Oxide Presen Range Average Samples Fe203 100 1.05 - 1.9 1.61 MgO 100.24 - 1.1.64 MnO 100.03 -.12 -07 Ti02 100.003 -.12.068 CaO 100.001 -.004.0026 SrO 0 BaO 100.001 -.007.004 Na20 100.65 - 1.0.86 Li20 40.04 -.05.018 Rb2O 0 Cs20 0 Ga203 100.01 -.022.018 Sc203 0 Co203 0 Cr203 60.002.0001 Sn02 100.024 -.10.058 V205 60.003 -.011.0004 F 80.14 -.21.14 * Based on spectrochemical data by C. E. Harvey 125

TABLE XXVI * MINOR AND TRACE ELEMENTS IN MUSCOVITES FROM PEGMATITES (Petaca District, New Mexico) Present in % of Oxide ~.Samples Range Average Fe203 100 3.2 - 5.6 4.24 MgO 100.008 -.4.119 MnO 100.28 -.83.47 Ti02 100.045.27.105 CaO 100.001 -.025.005 SrO 0 -- BaO 100.0004 -.06.0059 Na20 100.75 1.6 1.08 Li20 88.29 -.85.46 Rb2 0 O - Cs2O 0 Ga203 100.024 -.062.041 Sc205 6.004.0002 Co203 12.0004 -.0006.0001 Cr203 6.0002.00001 Sn02 100.003 -.04.014 V205 29.0005 - o002.0004 F 100.65 - 1.48 1.00 * Based on spectrochemical data by C. Eo Harvey 126

TABLE XXVII* MINOR AND TRACE ELEMNTS IN MUSCOVITES FROM PEGMATITES (Eight Mile Park District, Colorado) -- - 11 j-*-!-. —. -. -_ _ -. —--—, —-----------------. -—. —--—..1, -. --.- -- *- j,Oxide Present in of Range Average Samples Fe203 100 2.3 - 4.7 3.25 MgO 100.13 - 1.15.79 MnO 100.049 -.28.102 TiO2 100.008 -.31 094 CaO 100.001 -.005.0023 SrO 0 - BaO 69.0002 -.007.0011 Na20 100.62 - 1.8 1.12 Li20 69.06 -.52 17 Rb2O 0 - Cs2O 0 Ga2O3 100.014 -.048.025 Sc203 42.002 -.005.0016 Co203 16.0002 -.0003.00004 Cr203 10.001.0001 Sn02 53.001 -.056.006 V205 37.001 -.007.0015 F 100.10 - 1.15.48 * Based on spectrochemical data by C. Eo Harvey 127

TABLE XXVIII * MINOR AND TRACE ELEMENTS IN MUSCOVITES FROM PEGMATITES (Southwestern Montana) Oxide Present in % of Range Average Samples... Fe203 100 1.7 2.2 1.97 MgO 100.53 - 1.4.84 MnO 100.1.0 6.028 Ti02 100.01 - 2.12 057 CaO 100.00.001 -.004.00 SrO 17.003.0005 BaO 100.002 -.19.036 Na20 100.87 - 1.15 1.00 Li20 0 - Rb2O0 0 --- Cs20 0 Ga203 1PO.011 -.033.019 Sc203 17.003.0005 Co203 50.0003 -.0005.0002 Cr203 17.001.0002 SnO2 100.003 -.012.006 V205 83.0006 -.009.0004 F 17.10.017 * Based on spectrochemical data by C. E. Harvey 128

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN E. Commercial Quality Factors Factors that affect the utility and economic value of sheet muscovite are: (1) size of books, (2) structural defects, and (3) inclusions. The book size varies with the zone, deposit, and district and even greatly within a single zone. The relative abundance of structural defects such as A, herringbone, other reeve structures and wedge structure is related markedly to the type of occurrence of the muscovite, and governs the maximum size of flat, perfect sheets that may be cut from the books. The distribution of rulings and cracks also is, to a lesser extent, related to the zone type in which the muscovite occurs. These defects also limit the size of sheets obtainable from larger books. Pinholes, most of which result from the libertation of small included grains of crystals of various minerals during split — ting, are more abundant in ruby or brown micas of wall zones. Haircracks are likewise concentrated in this type. Both defects result in conductance through the mica sheet or film with attendant rupture or puncture, if the hole or-crack does not extend entirely through the sheet. Inclusions of hematite and magnetite ("staining") are more common in green than in ruby micas. Heavily stained mica, generally classed by visual inspection as electrical stock (No. 3), may contain a high proportion of grade E-l (Jahns and Lancaster 1950, p. 46). These investigators state (p. 46), "There appears to be a partial correlation between power factor and the amount: of hematite-magnetite stain in the mica, in that most of the E-2 and E-3 pieces are heavily stained but evidently there are numerous exceptions Some pieces that qualified as E-l, for example, were more densely stained than many with a higher power factor." Likewise there is no general correlation between power factor and color. Despite the antipathy of industry toward green mica, correctly prepared sheets and films of green muscovite are similar in quality to ruby muscovite. Possibly the aversion to green mica lies to some extent in the fact that, owing to a higher proportion of structural defects, large flat sheets are more difficult to prepare from it. 129

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN PART III,; L EPIDOLITE I.. MINERALOGY A. Composition The generalized formula for lepidolite may be written as: K2 (Li,A1 ) -6 ( S i 6A10 2-21-( F, O ) 3 In addition to the substitutions indicated by the formula, the following appreciable isomorphism has been noted: for K - Na,Rb,Cs; in octahedral positions -- Fe2,Fe3,Mn^Mg. The following elements have been reported and when present are usual in extremely small amounts: CaSn,GaCb,TiBa,Sr; Rb and Cs in some types are foundin considerable amounts (Rb2O = 5.2 percent and Cs20 = 1.90 percent, Lundblad, 1942); MnO in amounts as high as 6.0 percent reported in new spectrochemical analyses prepared for this report. The OH:F ratio varies considerably and OH may become negligible. Many unique theories have been advanced to explain the chemical variation in this species. The two most recent and widely accepted at present are those of Stevens (1938) and Winchell (1942). Both employ the end-member concept whose fundamental units are: Stevens (1938) KLi2AlSi4010F2 polylithionite K(Mn,Fe,Mg) AlSi010 (OHF )2 biotite 130

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN K4Li6A16A14Si12040 (OH,F)8 lithium muscovite KA2 (AlSi3)00 (OH,F) muscovite Winchell (1942) polylithionite K2Li4A12Si8020F4 protolithionite K2LiFe4Al3S i6020F4 paucilithionite K2Li3A15Si6020F4 Winchell (1942, p.121) points out that an additional end-member, muSeCvite, is needed to explain the lithia micas. However, he accepts only the above three as end-members of the lepidolite system and considers that the "muscovite molecule enters into crystal solution in this system only very little, if at all." He further notes that almost all lepidolites are deficient in Li20 by approximately 1 percent and considers this due to interlayered, but not isomorphously combined, muscovite. The end-members of Stevens (1958) and Winchell (1942) are useful today, from our point of view, inasmuch as they indicate the wide possible chemical variation occurring in lepidolites. The work of Stevens (1938) and Winchell (1942) is typical of those who favor the end-member concept. It has been pointed out by Heinrich (1946), Rabbitt (1948), and others that the theoretical end-molecules have no significance in the structure of the micas and other minerals and that variation in chemical compositions of various series cannot be adequately expressed by them. Collins (1942) further notes that the suitability of plotting optical properties against these hypothetical compounds is doubtful. For this study 26 lepidolites were analyzed spectrochemically. The results are shown in the Table XXIX. B. Structure 1. Polymorphism. Hendricks and Jefferson (1939) found three different polymorphic variations in lepidolites: 1-layer monoclinic: ao = 53.A bo = 9.2 c = 1 0.2A = 100 Space Group: Cm 131

TABLE XXIX* NEW SPECTROCHEMICAL DATA ON 26 LEPIDOLITES Oxide Present in Average Range Average of Samples Fe203 100.03-3.2.48 MgO 100.007-.27.043 MnO - 100.014-6.0 1.16 T1O2 35.005-.12.014 CaO 100.00-.001-.057 SrO 69.002-.02.006 BaO 65.0002-.001 o.0003 Na20r l100.14-.72.41 LiJQ 100 1.3-4.4 3.77 Rb20 100.12-1.0.54 Cs60" 62.04-.37.08 Ga203 100.003-.046. 017 Sc203 8.002- 013.0006 Co203 0 Cr203 4.0001..o000004 Spng0 100.002.16.021 V2o5 8.0005-.002.0001 F 100 1.9-6.8 5.41 *Based on spectrochemical analyses of C. E. Harvey 132

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN 6-layer monoclinic: a = 9.2= b = 5o3A = 60.OA g = 90~ Space Group: Cc This polymorph is unique in that the values of a and b are 0 O interchanged as compared with all other mica polymorphs. 3-layer hexagonal: ao = 5.35 co = 30.OA Space Group: C3112 or C3212 Approximately 300 samples (50 analyzed chemically) of lepidolite from about 100 different- specimens have been X-rayed and their symmetry has been determined. All the lepidolite polymorphs described by Hendricks and Jefferson (1939) have been recorded, and in general the results agree with theirs. One anomalous type discovered is a lepidolite whose structure is that of the normal 1-layer monoclinic polymorph but whose optic plane is normal to the side pinacoid instead of parallel to it. Unfortunately, owing to its small size, the crystal was lost during manipulation of the goniometer head upon which it was mounted. Also a new lepidolite polymorph has been found (3-layer monoclinic) which is described at the end of this section. Much of the structural data on lepidolites has been published in a paper, "Studies in the Mica Group; Relationship between Polymorphism and Composition in the Muscovite-Lepidolite Series," by Levinson (1953), which resulted from the natural-mica research supported by the Signal Corps contract. The information on the following pages represents an elaboration and extension of this information. 2. Single-Crystal Variations. Most of the specimens analyzed by Stevens (1938), which were structurally investigated by Hendricks and Jefferson (1939, p. 761), were restudied. Table XXX illustrates the high degree of cor — relation between Hendricks and Jefferson (1959) and the writers' data. The two sets of results are in general agreement. However, some very significant variations may be observed. In Stevens (1938) No. 3, two very small pieces of lepidolite were found from which Weissenberg patterns could be obtained. Both of these gave identical 6-layer monoclinic diffraction patterns. Inasmuch as an optical orientation method is mainly used, the mass of the sample could not be studied by the Weissenberg method because most of the crystals had wavy extinction and gave distorted interference figures. For Stevens No. 6, Hendricks and Jefferson (1939) report the 6-layer monoclinic structure. 13355

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Two Weissenberg photographs of this sample indicated respectively the presence both of the 6-layer lepidolite polymorph and the 2-layer lithian muscovite type. discussed previously. Therefore, the chemical analysis presented by Stevens (1938) is a composite analysis of the two forms. TABLE XXX STRUCTURE OF MICAS ANALYZED BY STEVENS (1938) Total No. of Stevens Atoms with Hendricks and This (1938) Li20 Six-fold Jefferson Report No__. _ Coordination (1939) 1 2.70 2.48 Muscovite Normal muscovite 2 3.51 2.54 Too fine grained 3 3.70 2.60 Too fine grained 6-layer monoclinic 4 3.81 2.58 for study Too fine grained 5 3.96 2.635 Too fine grained 6 5.04 2.73 6-layer monoclinic 6-layer monoclinic and lithian muscovite 7 5.05 2.86 6-layer monoclinic Not available 8 5.11 2.71 Single layer Single layer 9 5.33 2.81 Single layer Single layer 10 5.39 2.94 Single layer Not available 11 5.51 2.85 None available Not available 12 5.64 2.85 6-layer monoclinic Not available 13 5.78 2.91 Single layer Single layer 14 5.89 2.94 3-layer hexagonal 5-layer hexagonal 15 6.18 2.95 Single layer Single layer 16 6.84 2.98 Single layer Single layer 17 7.26 2.97 Single layer Single layer A systematic study was made of one book of lepidolite for the purpose of determining the extent of polymorphic variation within a single crystal. Specimen No. 514 from the Opportunity pegmatite, in Gunnison County, Colorado was chosen. The specimen is approximately 1-1/2 inches across and 3/4 inch thick. It has a uniform, typical pink color and is embedded in a mass of cleavelandite with grains of microlite. The sheets are a composite of many crystals. Between many of these crystal units are fine-grained aggregates of lepidolite with irregular form and anomalous extinction. The optic 134

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN planes of the various crystal units in a single sheet are always at 30~, or some multiple thereof, to each other. This appears to be the case regardless of the shape of the crystals in question, and whether or not they are separated by the fine-grained aggregate. 2V on all sections of the sheet, regardless of polymorph, remains almost invariant. In all, 27 Weissenberg photographs of small crystals from two sheets, one near the top and the other in the center of-'the book were made (see Table XXXI). Of the 14 photographs taken from the sheet near the top, 12 show the 6-layer monoclinic structure and 2 the 1-layer monoclinic structure. However, on 7 of the photographs of the 6-layer monoclinic form, the pattern of lithian muscovite was superimposed, in some cas ss only faintly present. Of the 13 photographs obtained from crystals in the:;en tral sheet, 12 had the 6-layer monoclinic structure and one had the 1-layer monoclinic pattern; only one lithian muscovite pattern was recorded and'this again was superimposed faintly on a 6-layer monoclinic pattern. This shows the wide polymorphic variation possible in lepidolite both within sheets and also parallel to the c-axis direction. Yet one polymorph, in this case the 6-layer type, is dominant throughout the book. An inspection of Table XXX indicates that except for numbers 12 and 14, those lepidolites with more than 5.1 percent Li20 have crystallized with the 1-layer structure. It is necessary therefore to explain the apparently erratic results obtained in the case for No. 12, which has the 6-layer structure and No. 14 with the 3-layer hexagonal structure. In the case of No. 12, unfortunately, no material was available for restudy. It seems possible however, that further investigation could reveal 1-layer forms present in addition to the 6-layer type found by Hendricks. A redetermination of the Li20 content also would be desirable. 3. Uniaxial Lepidolite. Uniaxial lepidolite (No. 14) which has crystallized with the 3-layer hexagonal structure has a composition identical with that of lepidolites that have crystallized as the 1-layer polymorph. Such uniaxial lepidolites or nearly uniaxial lepidolites are very rare. A review of the literature revealed that -Baumhauer (1903, 1912) first described such material from Mursinsk. He proposed the term microlepidolite for lepidolites of this character in contrast to macrolepidolites with large 2V. No chemical analyses were reported. Unixial lepidolite (5.83-6.15 percent Li 0) from Western Australia has been described by Winchell (1925), Simpson (1927), Murray and Chapman (1931), Stevens (1938) and Hendricks and Jefferson (1939). This uniaxial lepidolite is associated with a normal biaxial mica of almost identical composition. Winchell (1925, p. 424), in agreement with Simpson (1927), suggests that the uniaxial character "may be due to fine twinning on (001)." Hendricks and Jefferson (1939) reported the structure of uniaxial material from Londonderry as the 3-layer hexagonal type, whereas the biaxial material was the 1-layer polymorph. Jakob (1927) analyzed uniaxial lepidolite (4.93 percent Li20) from Usakos, and Pehrman (1945) analyzed and described similar material from Kimito (4.99 percent Li20). Specimens of both of these analyzed micas were found to have small areas of a biaxial phase. X-ray studies show that' the uniaxial 135

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN TABLE XXXI EXTENT OF POLYMORPHIC VARIATION IN ONE LEPIDOLITE BOOK FROM OPPORTUNITY PEGMATITE, GUNNISON COUNTY, COLORADO Sheet near top of book 514 (a) - 6-layer monoclinic (b),- 6-layer monoclinic + 2-layer lithian muscovite (faint) (c) 6-layer monoclinic + 2-layer lithian muscovite (d) 6-layer monoclinic + 2-layer lithian muscovite (e).- 6-layer monoclinic + 2-layer lithian muscovite (faint) (f) llayer monoclinic (g)- 6-layer monoclinic + 2-layer lithian muscovite (h)-, 6-layer monoclinic + 2-layer lithian muscovite (very faint) (i) — 6-layer monoclinic (j)-_ 6-layer monoclinic (k) 1-layer monoclinic (1) 6-layer monoclinic + 2-layer lithian muscovite (m) 6-layer monoclinic Sheet near center of book 514 (l). 6-layer monoclinic (2) 6-layer monoclinic (3) 6-layer monoclinic (4) 6-layer monoclinic (5) 6-layer monoclinic (6) — 6-layer monoclinic (7)-^- 6-layer monoclinic (8)- 6-layer monoclinic + 2-layer lithian muscovite (faint) (9) 6-layer monoclinic (10) — 6-layer monoclinic (11) 6-layer monoclinic (12) 6-layer monoclinic (13) 1-layer monoclinic Diffuse scattering observed in: 514 (a) 514 (1) (b) (6) (e) (7) (g) (8) (h) (10) (i) (j) 136

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN portions have crystallized as the 5-layer hexagonal polymorph but that the biaxial parts have the lithian muscovite structure. Since it has been demonstrated that the lithian muscovite structure is confined to a lower Li2O range the Li20 contents of the Usakos and Kimito micas represent composite results which are lower than those obtainable from strictly uniaxial specimens. Lundblad (1942, p. 58) notes that an analyzed specimen of lepidolite from Varutrask has both uniaxial and biaxial portions. The analysis of this material seems unreliable; 65.62 percent Si02 in a lepidolite is too high. Miss Berggren (the analyst), according to Lundblad (1942), seems to suspect quartz contamination. Under these conditions it seems best to eliminate this mica from consideration. Recently Macgregor (1945) described a lepidolite from Southern Rhodesia with a small 2V. A specimen similar to the described material surrounds a core of muscovite and in turn is enclosed by lepidolite with a large 2V. The muscovite structure is normal, and the lepidolite with the large 2V has crystallized as the 1-layer polymorph, whereas the uniaxial portion has crystallized as the 3-layer hexagonal form. The association of the 1-layer and 5-layer hexagonal polymorphs is identical with that of the Western Australia material. Chemical analyses of these lepidolites are not available. If one takes into account the presence of lithian muscovite in the Usakos and Kimito lepidolites, it seems reasonable to infer that the Li20 content of lepidolite of entirely uniaxial character may approach or even exceed the lowest Li20 content of any of the samples with the 1-layer structure. Structurally the 3-layer hexagonal polymorph can be geometrically interpreted as three 1-layer forms "twinned" on (001) with succeeding layers rotated 120~ in the same direction (either clockwise or counterclockwise). Hendricks and Jefferson (1939, p. 746) with relation to the 3-layer hexagonal structure state that, "a particular layer is repeated about a three-fold screw axis normal to the cleavage." The powder patterns (Fig. 7) of the 3layer hexagonal polymorph and the 1-layer monoclinic polymorph are almost identical. The cause of this rotation remains to be explained. Reusch (1869) first obtained an almost perfectly uniaxial figure in muscovite by stacking rotated sheets in this fashion. Although his sheets were of considerable thickness, the over-all optical effect appears to be the same. Ford (1932) explains the uniaxial effect on the basis of rotation of the plane of polarization. He notes (p. 328)n 137

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN "A particularly interesting case,. is that of the special properties of superposed cleavage-sections of mica. If three or more of these, say of rectangular form, be superposed and so placed that the lines of the axial planes make equal angles of 60~ (45~, etc.) with each other the effect is that polarized light which has passed through the center suffers circular polarization, with a rotation to the right or left according to the way in which the sections are built up. The interference-figure resembles that of a section of quartz cut normal to the axis.. Further, it is easy from this to understand how it is possible to have in sections of certain crystals (e.g., of clinochlore) portions which are biaxial and others that are uniaxial, the latter being due to an intimate twinning after this method of biaxial portions," Similar effects have been reported in feldspars (Sugi, 1940), talc and serpentine. 4. Powder X-ray Studies. Hendricks and Jefferson (1939) note that Stevens (1938) Nos. 2,3,4, and 5, which appear to be in the interval between the 2-layer muscovite type and the 6-layer lepidolite type, are too finegrained for study and speculate (p. 763): "It is tantalizing to think that samples 2 to 5 owe their poor crystal development to their close approach to the limit of the lepidolite solid solution in muscovite." It seems very likely that these fine-grained micas owe their macrostructural defects to small-scale variations in their crystal structure which itself probably is assignable to their chemically transitional position in the muscovite-lepidolite series. In order to test this possibility, powder X-ray studies on these micas were attempted. It was first necessary to obtain powder X-ray data for each of the known polymorphs. The only powder data on muscovites or lepidolites found in the literature were for normal muscovite (Nagelschmidt, 1937) and for the 3-layer muscovite polymorph (Axelrod and Grimaldi, 1949). Grim and Bradley (1951) list partial data for normal muscovite and a 1-layer lepidolite and give photographs of the two patterns. A complete set of powder pictures was compiled by X-raying powdered single crystals whose structure had first been determined by the Weissenberg method. This set of standard photos, which includes the polymorphs of muscovite and lepidolite under study, is shown in Fig. 7; the d-spacings of forms not previously published are given in Table XXXII. The structures of the fine-grained lepidolites analyzed by Stevens (1938) have been determined: 138

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Structure Total by Stevens Per Cent kThis Report SixNo. Lio 0 Hendricks --- No. i2o..... Fol.d i2 0 andand and Weissenberg Powder Coord, Jefferson 2 3.51 too fine- 6-layer lepidolite + 2.54 grained 2-layer muscovite 3 3.70 6-layer 6-layer lepidolite + 2.60 too fine- 2-layer muscovite grained 4 5.81 too fine- 6-layer lepidolite + 2,58 grained 2-layer muscovite 5 3.96 too fine- 6-layer lepidolite + 2.65 grained (2-layer muscovite?) That these fine-grained lepidolites are combinations of forms and not single structures confirms the idea that the poor crystal development is related to their composition. Micas with these combined structures may be termed transitional. It is noteworthy that crystals of Stevens No, 3 large enough for Weissenberg photographs are 6-layer forms, but the poorly developed crystals from the same specimen have a combination 6-layer lepidolite and 2layer muscovite structure. It is likely that the well developed crystals have a higher lithium content than the intergrown portions. In all cases the 6layer lepidolite polymorph predominates over the muscovite type in the transitional types. Owing to the close similarity between the powder patterns of normal muscovite and lithian muscovite, the exact type of muscovite in the combined powder patterns of the transitional micas is difficult to determine with certainty. However, careful study of the intensities indicates the probable presence of the lithian muscovite type. 5. Variation of Structure with Composition. On the basis of the structural data obtained it seemed advantageous to plot Li20 content against the polymorphic types of the muscovite-lepidolite series, normal muscovite, lithian muscovite, transitional micas, 6-layer lepidolite; 1-layer lepidolite and 5-layer hexagonal lepidolite. The result of plotting the Li20 contents of 55 analyzed micas against their structures is presented in Fig. 8. Many more micas with less than 2 percent Li20 were studied, but there is no need to record these, as they all have the normal muscovite structure. ------------------ 139

I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~x X:i:...............~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...i:::s:-:tS~~~~~~~~~~~~~~~~~~~~~~~~~~~ 1.Ix~~~~~~~~~~~~~~~~ X.T~i.1...... 1!~i!11~!i~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i......~~~~~~~~~~~~~~~~~~~~~~~~~.... Ii~~~~~~~~~~~~~~i: 1:I:::i:: iiriii~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~i~~~~~~~~~i~~~~~~~~~iiiriii iii lit~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~,,,iiiiiiiiiiiiii i!!i~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~- iii~: ~,!R: ~ii.....~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~........ i? ~::!iii~.b r!iiii!i 11 ~1 ~i i 3 Ii! I< a~~~~ Fig. 7. X-ray powder- photographs of polymorphs in muscovite-lepidolite series. A, normal muscovite, Varutra~'sk, Sweden; B, lithian muscovite, South Portland, Maine; C, 6-layer lepidolite, ~'Gunnison County, Colorado; D, l-layer lepidolite, Skuleboda, Sweden; E, -layer hexagonal lepidolite, Londonderry, Western Australia. (For another photograph see Levinson, 193 P 9::):::i:: 1 9 53: P::::::::::: 9:8::::::: -::.:::::::: S:::::0:::::::::: i::::::::: -- -::::::::::::: -: -::::::::: i I l i I 1 _I 1, 1 i iiiiuiiiiiiiii r lSl ii i fS Fig. 7. X-ray powder phot~~~~~~~~~~~~~~~~~~~~~~~~~~~~s~~~graphs of polymorphs in muscovite-lepidolite series. A, normal mus-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~1:1 pidolite, ondonderry Westerniii Ausraia (Fo anothri'ib phtorah se Leinon 1953 p. 98.- )::- -- _

TABLE XXXII SPACINGS OF POLYMORPHIC FORMS Cu Ka1, = 1o53736 kX 2-layer lithian. -layer lepidolite 6-layer lepidolite muscovite 1-layer lepidolite (hexagonal) (monoclinic) ((monoclinic) clon-doJ&arry, Gunnison South Portland, opsham, Western County, S o Maine Maine aine Australia Colorado No. 471 No0 467 No. 978 No. 514 I d-spacing I d-spacing I d-spacing I d-spacing m 9095 s 9.91 m 991 m 9ms 9.89 m 5.01 s 4.98 m 4,97 m 4.99 m 4,45 w 4,94 m 4066 m 4.49 vvw 3.95 vw 4,34 m 3.84 w 3584 vvw 3.87 vw 4,12 m 3.58 m 3.61 w 3,71 vw 3.86 vs 3.31 m 3.47 w 5.46 s 3 61 s 3 09 m 3.31 ms 3032 vs 3533 s 2,86 m 3o19 m 3521 s 3.07 w 2.651 m 3.07 m 2.98 s 2o87 s 2.571 m 2.88 s 2,84 m 2.675 mw 2 455 m 2 775 w 2.755 s 2.573 mw 2.375 vs 2,572 ms 2 571 m 2,468 vw 2.243 m 2,416 m 2.474 m 2 387 vw 2.186 vvw 2.248 m 2.387 vw 2,253 mw 2,126 vvw 2.190 vw 2.247 m 2o132 vw 2.052 vw 2.039 vvw 2.196 ms 1.988 m 1.984 s 1.985 mw 2.132 w 1o956 vvw 1o955 vvw 1 684 m 2,081 w 10748 w 1 718 vvw 10633 vw 1951 vw 1.715 mw 1 643 vvw 1 572 w 1,742 m 1,646 w 1o611 m 1o506 vw 1 o720 vvw 1 581 w 1 576 vvw 1.393 m 1 644 vvw 1.544 w l 547 vvw 1.355 vvw 1.596 m 1. 511w 1.511 vvw 1 319 vvw o 557 vw 1, 493 vvw 1.481 w 1.300 vvw 1.510 vvw 1o420 vvw 1.457 vvw 1o239 w 1o500 vvw 1o375 vvw 1 435 w 1.487 vvw 13552 vvw 1,411 vvw o453 vw 10337 w 1.341 vw 1.427 vw 1o299 vw 1.295 w 1.340 vvw 1.242 vvw 1.285 w 1.296 vvw 1.199 vvw 1o269 vvw 1.136 vw 1.243 vvw 1.220 vvw 1.199 141

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN The graph shows that micas with less than ca. 3.3 percent Li 0 have the normal muscovite structure; those with 3.4 to 4.0 percent Li2O have transitional structures; those with 4.0 to 5.1 percent generally have crystallized with the 6-layer lepidolite structure; whereas those biaxial micas with more than 5.1 percent Li20 generally have the 1-layer lepidolite structure. The uniaxial lepidolites as discussed previously have relatively large Li2O contents, probably in the same range with those of the 1-layer form, and this structure may be explained by "twinning." Too few data are now available to place the lithian muscovite type of strecture accurately in the series. Hpowever, the available evidence shows that it belongs somewhere between normal muscovite and 6-layer lepidolite. The above conclusions are graphically represented in Fig. 9o Analyzed micas described by Berggren (1940, 1941) and Lundblad (1942) do not fit well into the conclusions represented by Fig. 9. The results ob- tained from X-ray studies of these micas are presented in Table XXXIII. Inconsistencies appear with regard to Berggren's (1940) micas in A,B,C, and Lundblad's (1942) No. 10 and No. 13,Micas C and No. 13, which are reported to have 3.9 and 5.7 percent Li20, respectively, have crystallized with the normal muscovite structure. As much as 5.7 percent Li20, or even 3.9 percent Li20, in normal muscovite appears improbable on the basis of present knowledge. It is possible that the lithium content of the mica samples varies so much that the portions supplied represent extreme structural variations of the axalyzed materials. Several of the Swedish micas are optically nonhomogeneous, Vhich may indicate the presence of several crystal phases, and one may even be contaminated with quartz (Lundblad, 1942). Under such circumstances an attempt to correlate polymorphism with chemistry would be fruitless. Baumhauer (1912) described several lepidolites with a wide range of 2E and several distinctive types of etch figure in single specimens. Both of these features probably indicate structural variations. Except for the Swedish micas the correlation is good; indeed, better than might be expected, considering thewresence of structural variation in single books and the probable wide range in quality of the Li20 determinations. 6. Discussion. There are at least two possible ways in which a theoretical series between muscovite and lepidolite can be constructed. One substitution consists of 2Li(6) Si(4)_A1 6) A4); the other is ri(6)A1(4) Si(4) (the superscript numbers refer to coordination number). The former process leads to Winchell's (1942) polylithionite, whereas the latter leads to paucilithioniteo Minor complications arise with the substitution of Fe+2, Mn+2, and Mg for Al+3, but it is possible to determine without difficulty wihch of the above processes is involved, by inspection of the SiO2 content. The silica content will be higher than in normal muscovite in the first case and lower in the second. It is possible that the two substitutions operate together in natural micas, but there appears to be preponderant reason for believing that the sequence muscovite-polylithionite dominates, for all true lepidolites that we have been able to study are higher in SiO2 than muscovite. In the theoretical series muscovite-polylithionite I2Li(6)Si(4) 142 ----- ---- -- --

LEGEND FOR FIGURE 8 1, Wodgina, Western Australia; No. 36, Rowledge (1945). 2. Tabba Tabba, Western Australia; No. 52, Rowledge (1945). 3. Varutrask, Sweden; Analysis D, Berggren(1940). 4. Wodgina, Western Australia; No, 6, Rowledge (1945). 5o Tabba Tabba. ie-stiern Australia; No, 6, Rowledge (1945). 6. Karibib, South Africa, 2.5% Li20, courtesy of Foote Mineral Company. 7. Wodgina, Western Australia; No, 2, Rowledge (1945). 8. Manitoba, Canada; No. 1J Stevens (1938). 9. Karibib, South Africa; 3 o0% Li2O, courtesy of Foote Mineral Company. 10, Wodgina, Western Australia; Murray and Chapman (1931). 11. Tabba, Western Australia; Murray and Chapman (1931). 12 Ravensthorpe, Western Australia; Murray and Chapman (1931).; 13. Bear Claim, Manitoba; Spec. XXXIII, Ellsworth (1932)o 14. Pala, Califo.; No 2, Stevens (1938). 15o Pala, Calif.; No. 35 Stevens (1938), 16. San Diego Coo, Calif,; No, 4 Stevens (1938). 17. Pala, Calif.; No. 5 Stevens (1938) 18. Ubini, Western Australiaa Murray and Chapman (1931). 19, Poona, Western Australia; Murray and Chapman (1931). 20. Pala, Calif.o No. 6, Stevens (1938). 21, Ohio City, Colorado; No. 7, Stevens (1938)> 22. Pala, Calif.;. N. 12, Stevens (1938). 25. San Diego Co., Calif.; No. 8,Stevens (1958). 24. Mesa Grande, Calif.; No, 9, Stevens (1938). 25. Antsongombato, Madagascar; Mauguin (1928B), 26. Wakefield, Canada; No 10, Stevens (1938). 27. Mesa Grande, Calif.; No. 13, Stevens (1938). 28. Maharitra, Madagascar; Mauguin (1928B), 29. San Diego Co., Calif.; Noc 15, Stevens (1938). 30. Antsongombato, Madagascar; No. 16, Stevens (1938). 31. Greenland; No. 17, Stevens (1938). 52. Usakos, Southwest Africa; Jakob (1927). 55. Kimito, Finland; Pehrman (1945). 34. Calgoorie, Western Australia; No. 14, Stevens (1938)o 35. Londonderry,'Western Australia; Murray and Chapman (1931) The structures of all the above micas, with the exception of numbers 21,:22 and 26, have been determined by the authors* In the case of the three exceptions the structures reported by Hendricks and Jefferson (1939) have been used. 145

3-LAYER LEPIDOLITE 32 W3 I-LAYER LEPIDOLITE ) @ ( 0 6-LAYER LEPIDOLITE' @ 0 TRANSITIONAL ~ I LITHIAN MUSCOVITE NORMAL MUSCOVITE s~ 00 10 20 3.0 40 50 60 7.0 % L,2 - Fig. 8. Plot of the Li20 contents of 35 micas in muscovite-lepidolite series against the various polymorphs. 3-LAYER LEPIOOLITE I-LAYER LEPiDOLITE 6-LAYER LEPIDOUTE TRANSITIONAL UTHIAN MUSCOVITE NORMAL MUSCOVITE s' 0.0.0 1.0 2.0 3.0 4.0 5 0 6.0 7.0 Fig. 9. Idealized representation of relationship between Li20 content and polymorphism in the muscovite-lepidolite series.

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Al(6)A1 (4 which appears to be represented in nature, an increase in the Si content is accompanied by an increase of Li. In order for the overall charges to be balanced, two Li ions are needed in six-fold positions for each replaced aluminum. It is concluded, therefore, that inasmuch as two Li ions cannot occupy the space of one Al ion, the additional Li ion must occupy one of the vacant octahedral positions in the muscovite structure. The higher the Li content (i.e., the closer the approach to polylithionite), the more vacant positions are occupied; thus there exists a structural (though discontinuous) as well as a chemical series between heptaphyllite muscovite and octophyllite lepidolite.* TABLE XXXIII STRUCTURE OF MICAS DESCRIBED BY BERGGREN (1940, 1941) AND LUNDBLAD (1942) Analysis % Li20 Structure Structure A 5.95 6-layer lepidolite Varutrask B 4.35 1-layer lepidolite Varutrask C 3.9 normal muscovite Varutrask D 2.45 normal muscovite Varutrask E 1.80 normal muscovite Varutrask G 0.75 normal muscovite Varutrask H 0.69 normal muscovite Varutrask I 0.22 normal muscovite Varutrask J 0.76 normal muscovite Varutrask K 1.10 normal muscovite Varutrask L 1.1 normal muscovite Varutrask No. 10 4.55 1-layer lepidolite Varutrask No. 13 5.7 normal muscovite Uto No. 14 5.5 6-layer and 1-layer Rozena lepidolite Further evidence for the existence of the muscovite-lepidolite series is found in: Actually the term lepidolite is awkward because it embraces 3 different polymorphs. The intended meaning of lepidolite in the phrase "muscovitelepidolite series" is that of the ideal Li-rich mica with the composition of polylithionite (KLi2AlSi401F2) whose octahedral positions are completely filled. 145

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 1. The interpretation of the structure of lithian muscovite as a less distorted normal muscovite structure owing to a filling of the structural holes by Li ions (see discussion of lithian muscovite). 2. The observation that the Li2O content of micas (Fig. 8) in the muscovite-lepidolite series ranges from 0 to 7.26 percent and that therefore the above mentioned accompanying substitutions [2Li(6)Si(4) F= A1(6)A1(4) probably take place. 3. The interpretation of Winchell's (1942) observation(that most lepidolites are deficient in Li20)as indicating that these micas do not have complete occupation of the vacant octahedral positions by Li ions; no muscovite has been found (by X-ray methods) interlayered with lepidolite except some lithian muscovite. 4. Stevens' (1938, p. 618) and Hendricks and Jeffersons' (1939, p. 761) calculations showing that the number of atoms with octahedral coordination in the micas analyzed by Stevens (1938) range from 2.48 to 2.98 (see Table XXX ). In our discussions of biotite and phlogopite it will be pointed out that no such similar chemical or structural series exists between heptaphyllite muscovite and either octophyllite biotite or phlogopite, as has been suggested by various investigators. It has been suggested that the evidence presented above also may indicate that the extent of solid solution in the muscovite-lepidolite series is very limited. The transitional micas that appear near the middle of the series might be composed of two phases in such proportions that their average composition contains about 2.55 atoms with octahedral coordination. This concept is difficult to accept owing to the fact that several micas close to the middle of the series, such as Stevens (1938) No. 1 which contains 2.48 atoms in octahedral coordination, contain, from X-ray and optical studies, only one uniform phase. 7. Summary. The discovery of the lithian muscovite type of structure, the interpretation of the structural origin of the 5-layer hexagonal lepidolite polymorph and good correlation between Li20 content and the various polymorphs shows that chemical and structural transitions occur between Lifree muscovites and Li-rich lepidoliteso It seems probable, however, that the relationship between a single-ion variation and polymorphism in a structure as complex as the micas is somewhat oversimplified. Overlapping undoubtedly takes place and it may be possible to get even more accurate results by calculating the number of atoms with octahedral coordination against the ________________________ 146 ___________

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN various polymorphs; at present too few complete analyses and specimens are available to attempt this approach. The study shows also that structural variations in single lepidolite crystals are common and can be coupled with compositional variations. The correlation may prove valuable in estimating the approximate Li20 content of the mica without a chemical analysis, Normal muscovite may have as much as 35. percent Li20. Micas with 3.4 to 4.0 percent Li20 are generally characterized by Coor erystal development and anomalous optical properties owing to their transitional structures, Lepidolites with 4.0 to 5.1 percent Li20 generally crystallize as the 6-layer monoclinic polymorph whereas those with more than 5.1 percent Li20 have usually crystallized as the 1-layer monoclinic polymorph. Hexagonal lepidolite is commonly associated with 1-layer monoclinic polymorph, and both have essentially identical compositions. C. A New 3-Layer Monoclinic Lepidolite Polymorph A specimen of lepidolite from Skuleboda, Sweden (No. 476) was found to include a very small amount of uniaxial (2V as large as approximately 5~) material. The majority of the lepidolite is biaxial (2V = 37~) and has crystallized as the 1-layer polymorph. The biaxial and uniaxial portions occur in the different sections of the same sheets and are megascopically indistinguishable. The boundary between the two phases is microscopically sharp. X-ray studies of the uniaxial portion reveal the presence of an unreported 3-layer monoclinic polymorph. Zero-level Weissenberg photographs have been taken about all aand pseudo a-axes" (120~ from true a-axis) as well as b- and pseudo b-axes of this polymorph. A plane of symmetry in the position of b* as well as c* was recorded on only one of the photographs; therefore this pattern is considered as having been obtained from rotation about the true a-axis. Coupled with information from other Weissenberg photographs the crystal system is established as monoclinic. The Weissenberg pattern obtained by rotation about the O-level a-axis is identical with 0-level a-axis photographs obtained from the 3-layer hexagonal polymorph. A O-level pseudo a-axis Weissenberg is shown in Fig. 10. Among others, 0-level b-axis (Fig. 11) and 1-level a-axis (Fig. 12) photographs show some differences from the corresponding patterns of the 3-layer hexagonal polymorph. Differences along the 135 reciprocal lattice line between the 3-layer monoclinic lepidolite and 35-layer hexagonal polymorphs may be seen by comparing Fig. 12 with the 1-level a-axis Weissenberg photographs illustrated by Axelrod and Grimaldi (1949). For all practical purposes the illustrations given by Axelrod and Grimaldi (1949) are identical with those which can be obtained from the 3-layer hexagonal polymorph originally described by Hendricks and Jefferson (1939). Similar differences are also obtained comparing any reciprocal lattice row in which k is divisible by 147

obtained from 3-layer monoclinic lepidolite.

2o~. Fig. 11. 0-level b-axis We issenberg photograph obtained from F-1ayer monoclinia 1epidolit~. ~~~~~~~~~~~~~~~~~~~~:..........-............. ~.......................................-..............: ~.......... Fig. 12. 1-level a-axis We is senberg photograph obtained from ~-1ayer meonoelinic lepidolite.

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 3 (hkji k = 3nj including hO reflections). Indexing showed that all reflections with h + k odd and (00Q) with 3 ~ 5n were absent. The possible monoclinic space groups with the information thus far obtained are Cm, C2, and C2/m. In comparing a- and b-axis photographs of the new form and the 1layer monoclinic lepidolite (space group Cm) a very interesting relationship is observedo When the photographs are superimposed, every third reflection of the new form corresponds exactly with a reflection of the 1-layer polymorph. In between, two additional reflections will almost always be found on photographs of the new polymorph. This relationship also applies to photographs obtained by rotation about the pseudo a- and b-axes as wello Therefore it is logical to assume that the 5-layer monoclinic lepidolite contains the same symmetry elements as the 1-layer polymorph and thus has also crystallized in the monoclinic space group Cm, The unit cell dimensions are (approximately) a = 5.2 A 0 bo = 9.0 A c0 = 30.0 k = - 90~ (+30') X-ray powder data of the 3-layer monoclinic lepidolite are given in Table XXXITV In comparing the d-spacings ad intensities with those given for the other mica polymorphs (Table XXXII) the existence of the new polymorph is substantiated. Also in Table XXXIV a quantitative spectrochemical analysis is given of the biaxial lepidolite (l-layer) which is found in the same book with the new structural variety. The Li20 content of this specimen, however, does not correlate well with the results obtained for other micas in the muscovite-lepidolite series. The relationship of the new 3-layer monoclinic polymorph to other members of the muscovite-lepidolite series remains to be deciphered. It appears, however, that this polymorph is extremely rare. D. Optical Properties Of Lepidolite 1. Indices, 2V, and Variation and Composition a = 1.525 - 1.548 2V = 0 - 58~ (negative) p = 1.551 - 1.580 CnA = 0 -3 0 = 1o554 - 1.586 PAa = 0 - ~ - a = 0,018- o0538 T = b Optic plane normally parallel to (010). 150

TABLE. XXXIV SPECTROCHEMICAL AND X-RAY DATA ON LEPIDOLITE FROM SKULEBODA, SWEDEN (No. 476) Analysis of 1-layer X-ray data on 3-layer lepidolite* monoclinic lepidolite** SiO2 = 48. Cu Kli; = 1.53736kX A120 = 50. Intensity d&-spacing K 20 = 11.1 s ---- - 9.91 m. —-.-.4.98 Fe203 =.22 m ----- 4.54 w 3 587 MgO =.01 m -----.56 S rr~~~iil 3.32 NaO2, =.32 w. —--- 3.18 -vw - 5. —- 3.11 MhO) = 1.1 vw ----. 2.95 vw ----- 2.87 BaD.: =.0008 vw ----- 2.83 ms ---- 2.60 Rb2Q =,80 vw ----- 2.54 w --- 2.44 L~0'Q = 4.1 ww' ----- 2.33 vw -.-. 2.26 SiO-2. =.006 vw -. —-- 2.17 v -, 2.09 Ga203 =.016 vvw ---- 2.02 T1i02. =.005 vw — 1.71 w --. —-.- 1.67 SC203 = --- vw — 1.63 vw ----- 1.60 SrO:, =.006 vvw ----- 1955 w --... 1.51 Cr203 = -- vw ----- 1.42 vvw- - 1.55 CaO' =.002 Vv:; - 1.30 vvw ---- 1.28 V25 =..002 vw ------ 1.24 Cs2Q' =.12 F. = 6. 5 **Measurements by C. H. Hewitt *Spectrochemical analysis by C. E. Harvey " 151

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN The relationship between variation in optical properties and the chemical composition is complex, but the values of the indices of refraction do not depend upon the content of Li, but rather upon the amounts of ferric iron, ferrous iron and manganese that may be present. Lepidolites that are low in these three elements have indices of refraction and birefringences in the lower portions in the above listed ranges, whereas lepidolites containing appreciable amounts of ferric iron particularly, or of some ferric iron plus ferrous iron and manganese, have indices and birefringences that lie within the upper part of the above listed ranges. Apparently there is also a slight decrease in the size of 2V with increasing ferric iron or with increasing ferric iron plus ferrous iron and manganese. Optical constants were measured on individual flakes whose structure first had been determined by the Weissenberg method. The results are presented in Table XXXV. The indices of refraction were measured on the Abbe refractometer or were obtained by the immersion method and are reproducible to + 0.001. The Mallard method was used in the determination of 2E and 2V was calculated. The only other determinations of the optical properties of lepidolite polymorphs have been by Hendricks and Jefferson (1959). Their results, though scanty, are in close agreement with those of this paper and are presented below. The values are averages and the numbers in parentheses indicate the number of published measurements used in calculating the averages. 6-layer lepidolite a = not reported P = not reported Y= 1.559 (3) 2V = 32~ (3) 3-layer hexagonal lepidolite a = 1.525 (1) D = not reported = 1.558 (1) 2V = 0~ (1) 1-layer lepidolite = 1.55533 (1) D = not reported P= 1.559 (12) 2V = 41~ (12) The lepidolite polymorphs are not distinguishable on the basis of optical properties alone; it follows therefore that their general chemistry cannot be deduced from the optical constants. A trend is that the indices, and particularly 2V, of the 1-layer polymorph tend to be slightly higher than those of the 6-layer polymorph, but the overlap is broad. This can be accounted for by the optical inactivity of the Li-ion. Although variations in Li content influence the layer stacking pattern, changes in optical constants are more sensitive to variations in Fe3,Ti,Fe2 and Mn, both in muscovite and the lepidolites. The indices of lepidolite vary within books (note values obtained from spec. 514 and 452 in Table XXXV). However, the indices of lithian muscovite are in the range of normal lepidolites whose values are considerably 152

TABLE XXXV INDICES OF REFRACTION AND 2V OF LEPIDOLITES AND LITHIAN MUSCOVITE Specimen Loca y 2V Number Locality 2V 6-layer lepi4p -it 502 Opportunity Pegmatite, Ohio City, Colorado 1.531 1.555 1.560 30~ 514(2) Opportunity Pegmatite, Ohio City, Colorado 1.531 1.553 1.557 30~ 514(5) Opportunity Pegmatite, Ohio City, Colorado 1.533 1.555 1.560 33~ 514(4) Opportunity Pegmatite, Ohio City, Colorado 1.533 1.557 1.560 33~ 514(5) Opportunity Pegmatite, Ohio City, Colorado 1.531 1.555 1.558 -- 514(6) Opportunity Pegmatite, Ohio City, Colorado 1.531 1.553 1.560 33~ 505(b) Brown Derby Pegmatite, Ohio City, Colorado 1.530 1.556 1.560 34~ 967(a) Stewart Mine, Pala, California 1.530 1.552 1.559 335 967(b) Stewart Mine, Pala, California 1.529 1.554 1.558 340 970(a) Stewart Mine, Pala, California 1.534 1.555 1.559 33~ 452(a) Varutrask, Sweden 1.532 1.556 1.561 33~ 452(b) Varutrask, Sweden 1.535 1.556 1.561 35~ 452(c) Varutrask, Sweden 1.532 1.552 1.557 37~ 35-ayqr hexagonl leepidolite 539(f) Pope mining claim, Southern Rhodesia - 1.568 1.568 0~ 5539(g) Pope mining claim, Southern Rhodesia - 1.566 1.566 0~ 1-layer lepidolite 514(k) Opportunity Pe.gmatite, Ohio City, Colorado - -1.556 1.558 - 464(b) Newry, Maine - 1.556 1.558 50~ 471(b) South Portland, Maine 1.533 1.553 1,560 41~ 476(b) Skuleboda, Sweden - 1.562 1.564 476(c) Skuleboda, Sweden - 1.564 1.566 37~ 535(c) Dogon Daji, Nigeria - 1.557 1.562 47~ 539(a) Pope Mining Claim, Southern Rhodesia - 1.562 1.565 380 539(b) Pope Mining Claim, Southern Rhodesia - 1.562 1.566 37~ 2-layer lithian muscovite IT 556(b) Tordal, Norway - 1. 13.556 34 556(c) Tordal, Norway 1.532 1.554 1.558 340 465(a) Newry, Maine 1.533 1.555 1.559 34~ 465(f) Newry, Maine 1.534 1.555 1.561 37~ 471(a) South Portland, Maine 1.529 1.551 1.556 32~ 679(a) Usakos, S. W. Africa 1.534 1.559 1.566 - Specimens numbered such as 514(2) through i14(6)y- or 452(a) through 452(c) represent measurements obtained from different flakes in the same book. 153

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN below those of normal muscovite. It is well known that indices are very sensitive to a small amount of certain isomorphous substitutions. Winchell (1942) showed that practically all lithia micas show variations of the indices within single flakes but attributed the phenomena to other causes. 2. Variation of Color with Composition. It has been demonstrated by Heinrich and Levinson (1955) that the rose muscovites owe their color to the essential absence of Fe2 and the dominance of Mn over Fe3. Under such conditions it is possible for the Mn to be oxidized to the higher valence (Mn3) in which state it becomes a powerful pigmenting agent, even though, present in only very small amounts. They state, (p. 46): "In lepidolites, the color variations also are related to fluctuations in the Fe/Mn ratio. The role of Mn in the coloration of lepidolite has been stressed, for example, by Hintze, 1897 and by Doelter, 1917. As stated by Shibata (1952, p. 160),'Lepidolite is pink or violet, which is due to the color of manganese. Lepidolite is sometimes colorless when it is poor in MnO. Lithium is not a color agent....' Those lepidolites that are colorless or of very pale colors usually have Fe:Mn = 1:1; those that are gray normally have Fe > Mn, whereas the purple lepidolites contain more Mn than Fe." In our work we have been able to confirm the above conclusions in greater detail on the basis of new spectrochemical analyses of lepidolites. From the data presented in Table XXXVI,XXXVII, and XXXVIII, the following generalizations can be drawn: Mn:Fe (total Fe) greater thank 6:1 deep shades of purple between 3:1 and 5:1 medium shades of purple between -1:1 and 3:1 colorless to pale tints of purple between.1:1 and 2:1 colorless to gray (where amount of Fe is less than lpercent) Where the Fe content rises much above 1 percent, it apparently masks the effect of equal or even larger amounts of Mn and results in a brownish color (see specimens 1143,1016,990 in Tables XXXVII and XXXVIII). The effect of small amounts of Ti seems considerable in some cases. Ti apparently adds a brown tint in specimens No. 504 and No. 514 (Table XXXVI). Accurate color descriptions of lepidolites as well as other micas are difficult to obtain. Physical factors which influence color are size of books, impurities, state of aggregation and grain size. The colors listed in Tables XXXVI, XXXVII, and XXXVIII are as accurate as we are able to ascertain. The specimens were compared under the same light conditions. It is instructive 154

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN to compare our color descriptions with those of Stevens (1938) on his analyzed micas made available for our study (Table XXXVIII). It would also be constructive to compare the variations in chemical analyses of selected oxides as listed in Table XXXVII. 155

TABLE XXXVI RELATIONSHIP BETWEEN COLOR AND COMPOSITION IN LEPIDOLITES; NEW ANALYSES* Number Locality MnO Fe203 TiO2 Color Remarks 464 a Newry, Me. o10.06 -- pale purple coarse-grained f same 464 b *.11.10 -- fine-grained pecimen 476 Skuleboda, 1,1.22.005 deep purple Sweden 483 Newry, Me..09.20 -- gray o\'488 - Pala, Calif..10.03 -- pale purple 4c99 Brown Derby,.23.05.015 medium purple Colorado 500 Bazooka peg, o.13 09 -- pale purple Colorado 503 Brown Derby,.45.o6.012 deep purple Colorado 504 Brown Derby, 1.7.17 o03 deep purplish Colorado brown *Spectrochemical analyses by C. E. Harvey

TABLE XXXVI (conto) Number Locality MnO Fe203 TiO2 Color Remarks 508 Brown Derby,,08.06 pale purple Colorado 511 Brown Derby, 005.05 -- very pale purple curved Colorado 514 a Brown Derby, 1.2.27.005 medium brownish Colorado purple two analyses of same spec514 b Brown Derby, 6,0.60 o01 medium brownish imen -\k Colorado purple 535 Nigeria.46 o04 - deep purple 539 a S. Rhodesia 5.5 1.1 o045 deep purple biaxial zone 539 c S. Rhodesia 6.0 2.2 o08 deep purple uniaxial zone 991 Auburn,.39.69 - colorless Maine 1097 Rozena.37.08 - deep purple

TABLE XXXVII RELATIONSHIP BETWEEN COLOR AND COMPOSITION IN LEPIDOLITES; REANALYSIS OF LEPIDOLITES REPORTED IN THE LITERATURE. Number Locality MnO Fe20 TiO2 Color Reference Remarks 521 Varutrgsk.55.10 medium purple this report -- Li2O = 5.8 521 Varutrgsk.28.10 0.00 Berggren (1940) fLi20 = 5.9 Anal. C LFeO = 0.00 551 Varutrask.10.04 gray this report — Li2O = 4,1 551 Varutrask.00.17 0.00 gray-white Lundblad (1942) fLi2O = 4,55 Anal. 10 L~eC = 0.00 552 Uto.18.18 - white this report —-- Li2O = 4,2 552 Uto.28.15 0.00 --- Lundblad (1942) fLio = 5.7 H Anal. 15 eO = 0.02 \J1 55533 Rozena.46.08 -- deep purple this report —-- Li2O = 2,1 55533 Rozena.52.06 0.00 --- Lundblad (1942) fLi2O = 55 Anal. 14 (FeO = 0.00 1145 Quebec 2.0 2.5,.12 pale brown this report —— LI2O = 4.2 1145 Quebec 4.19 2.52 -- -- Ellsworth, —-— Li2O = 5.44 (1952) Spec. XXXII "Spectrochemical analyses by Charles E. Harvey

TABLE XXXVIII RELATIONSHIP BETWEEN COLOR AMD COMPOSITION IN LEPIDOLITES; SPECIMENS IN OUR LABORATORY FOR WHICH ANALYSES ARE RECORDED IN THE LITERATURE Number Locality — MnO FeO Fe20O3 TiO2 Color Reference' 966 Ramona 2.06.21*.02 pink Stevens (1938) No. 15 967 Pala, Calif..92.04 tro medium do. No. 3 purple 969 Antsongombato.72.12.03 medium do. No. 16 purple 970 Pala, Calif..52.16 tr. very pale do. No. 6 purple 971 Mesa Grande.14.11 tr. pale do. No. 8 purple 972 Mesa Grande o17 o04 tr. medium do. No. 9 purple 973 Mesa Grande.61.07.01 medium do. No. 13 purple 975 Pala, Calif..65.05 tr. medium do. No. 5 purple 977 San Diego Co..28.04 tr. medium do. No. 4 purple 978 W. Australia.76.23 o06 pale do. No. 14 purple 990 Usakos 1.48 0.00 2.34 tr. pale Jakob (1927) browvn 1016 Kimito.75 1.52 1.55 0.00 pale Pehrman (1945) brown 1259 Maharitra.45 --.03 -- deep Maiguin (1928B) purple 1260 Antsongombato.24 --.52 colorless Mauguin (1928B) All Fe is reported as FeO by Stevens (1938) in the micas listed in this table. 159

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN II. OCCURRENCE A. Noepegmatite There are very few authenticated occurrences of lepidolites outside of pegmatites. The presence of lepidolites in the hydrothermal contact aureole of granites has been suggested by Harker (1932, p. 123). Lepidolite also has been reported from a few granites, aplites, and high temperature veins and lodes, particularly those containing tin and tungsten minerals. That these micas are lithium-bearing has been shown in some cases, but few have been demonstrated to be lepidolites in composition or in structure. B. Pegmatite. 1. Localizationo In pegmatites, lepidolite and lithian muscovites occur in: a. Fracture-fillings, not uncommonly with cleavelandite, quartz, and tourmaline. b. Replacement bodies, also commonly with quartz, albite (incl. cleavelandite), and a variety of accessory constituents. Some investigators believe that lepidolite also can form as a primary constituent of inner zones (Cameron, et al., 1949, p. 68), but the number of such examples, even if they are truly of magmatic origin, is doubtless small indeed. The major lepidolite areas of the world are: United States: Maine - Black Mtn., Rumford and Mt. Mica; Paris, Me. Colorado - Brown Derby, Gunnison County New Mexico - Harding Mine, Taos County California - Stewart Mine, Pala 160

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Southeastern Manitoba, Canada Varutrask, Sweden Haut Vienne, France Karibib, South-west Africa Pilbara area, Western Australia Madagascar Portuguese East Africa Southern Rhodesia 2. Macrostructural Features. a. General: Within individual pegmatites the grain size of lepidolite may vary enormously; although usually within a single replacement body, the grain size is relatively uniform. The grain size of lepidolite ranges from microscopic, which results in megascopically massive or structureless aggregates, to very coarse books, 10 to 12 inches across. Some of these larger books display structural defects identical with those found in muscovite-A, herringbone, wedge, and concentric structure. Mineral inclusions also may be common, as are fractures and rulings. Wavy and rippled books are very common, and even small books with completely flat sheets are rare. b. Overgrowths: Overgrowths of lepidolite on muscovite have been found in pegmatites at several localities. The lepidolite commonly has formed as a narrow rim around a broad core of muscovite. The contact between the two may be sharp and show the general euhedral crystal outline of the muscovite core, or it may be irregular and appear to be somewhat corroded. In one type a narrow intermediate band of lepidolite in parallel position and having a common basal cleavage plane with the muscovite is found in direct contact with the muscovite and a third zone of granular lepidolite forms the extreme margin of the crystal. This latter type of lepidolite is highly twinned in tiny rhombic units. In a few cases pink fibrous muscovite coats the exterior of the crystal (Bowman, 1902). Under the microscope, the twinned lepidolite aggregate does not extinguish as a unit. Because of the small size of the lepidolite units it is almost impossible to obtain the relationship of percussion figures and optic planes between the two species near the boundary. However, three successful determinations were made that show the optic planes of the two species to be either parallel or at an angle of 30~. The results of a study of a group of specimens of lepidolite overgrown on muscovite are given in the following list: ____________________________________ 161

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Specimen No. Location Notes 417 Eight Mile Park, granular lepidolite rim, straight Colorado boundary 461 Auburn, Maine single-crystal lepidolite rim, common basal cleavage, border irregular and corroded, optic planes parallel 467 Topsham, Maine granular lepidolite rim, boundary straight in some sections and corroded in others 468 Topsham, Maine granular lepidolite rim, boundary corroded 469 Topsham, Maine granular lepidolite rim, boundary corroded 470 Topsham, Maine granular lepidolite rim, boundary corroded 539 Southern Rhodesia broad sheet of muscovite with inner rim of uniaxial muscovite and outer rim of biaxial lepidolite, all three zones have common cleavage, optic planes at 30~ 651 Auburn, Maine granular lepidolite rim 652 Auburn, Maine single-crystal lepidolite rim, boundary straight, common basal cleavage 656 Auburn, Maine single-crystal rim of lepidolite, boundary straight, common basal cleavage, optic planes at angle of 30~ 713 Topsham, Maine muscovite core, thin rim of single crystal lepidolite, outside rim of granular lepidolite, boundaries corroded, muscovite and thin rim of lepidolite have common cleavage 717 Auburn, Maine granular lepidolite rim, corroded boundary Reports of overgrowths from Maine have been made by Clarke (1888), Wolff and Palache (1902), Bastin (1911), and Baumhuer (19125. Overgrowths have been found at Paris, Auburn, Mount Apatite, and Minot. Generally, the lepidolite is fibrous or granular around a core of muscovite. A few occurrences of crystallographically continuous lepidolite on muscovite have been reported from the Wade and Pulsifer gem quarries, Androscoggin County, Maine. One diamond-shaped book of muscovite a foot across with a border zone of lepidolite 4 inches wide was found there. 162

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN The occurrences at Haddam Neck, Connecticut,- have been studied by Bowman (1902) and Sterrett (1923). The columnar mica crystals occur with smoky quartz, albite, microcline, cookeite, and tourmaline. The outer surface of the mica crystals has a fibrous appearance caused by a thin layer of fibrous pink muscovite. The fibers are parallel to the long axis of the crystal. Cleavage in the fibers is perpendicular to their long axis. The lilac lepidolite surrounds a rhombic or hexagonal core of green-white muscovite. All three zones have a continuous cleavage. Under the microscope, the lepidolite does not extinguish completely because of the superposition of layers of material in twin position. Jahns and Wright (1951) report lepidolite rims fringing a green core of muscovite. The two micas are crystallographically continuous and have a common basal cleavage. Heinrich (1948) has observed similar overgrowths in the Meyers Quarry pegmatite, Eight Mile Park, Colorado. A regular overgrowth of lepidolite on muscovite has been reported by Scharizer (1886, 1887B) from Schuttenhofen, Germany. A rhombic core of muscovite is surrounded by a broad crystallographical continuous rim of pink lepidolite. Three separate irregular units were found in one specimen. In two of these units the optic plane is normal to the optic plane of the muscovite; in the third unit, the optic plane makes an angle of 30~ with the optic plane of the muscovite. Pehrman (1945) has described an interesting and rather complex overgrowth relationship between muscovite and lepidolite among the micas from Kimito, Finland. Several zones of lepidolite are involved. Overgrowths of lepidolite on biotite are rare and have been reported from only a few localities; no reports of optical relationships have been found. Landes (1925) describes a 6-mm-wide zone of lepidolite around biotite plates from gneiss xenoliths in the granite pegmatites of central Maine. He attributes the lepidolite to reaction between the biotitexenoliths and the lithia-rich magma. Clarke (1888) reports an overgrowth from Cape Ann, Massachusetts, in which a crystal of black lepidomelane is surrounded in part by a border of small crystals of a "dark greenish-black lithia mica, presumably cryophyllite". Analyses of the two micas are as follows: Border Core Granular Cryophyllite Lepidomelane Si02 52.17 31.69 Al20 16.39 11.93 Fe320 4.11 8.06 FeO 6.08 350.35 MnO 0.32 0.21 163

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Ca0 Trace 0.23 MgO Trace 0.05 Li20.453 Trace Na20 0.60 1,54 K2 10.54 8.46 H20 1.43 4,25 F 7.02 Trace TiO2 -- 3.42 3. Compositional Variation in Individual Pegmatites and Districts. Table XXXIX lists five new analyses of paragenetically distinct lepidolites from the Brown Derby pegmatite, Gunnison County, Colorado. Despite the fact that they differ markedly in grain size, texture, association and somewhat in color, the compositional differences are relatively small. The major differences occur in MnNa,H20+ and F; minor in Li. Analyses 7 and 10 are by two other analysts of our lepidolite 504. Table XL lists new spectrochemical analyses of tlhre lepidolites and one lithian muscovite from Newry, Maine; again the principle differences are in MnNa,F and Rb. Table XLI lists reanalyses of lepidolites from the Varutrask pegmatite, Sweden, analyzed by Berggren; the agreement between the two sets of results is good. In Table XLII are reported three analyses of lepidolites from two other pegmatites in the same district as the Brown Derby. Numbers 514a and 514b, from the same crystal, differ considerably from one another (notice MnO content) and also from 500, as well as from the Brown Derby lepidolites (Table XXXIX). Thus it appears that paragenetically distinct lepidolites may differ considerably in composition or in other cases they may be rather similar. 164

TABLE XXXIX ANALYSES OF IEPIDOLITES FROM BROWN DERBY PEGMATITE, GUNNISON COUNTY, COLORADO xide I... _.....Specimen Nos.* 499 03 504 508 511 7 10 SiO2 49.97 51.99 49,88 54.82 50.50 49.58 49.19 A1203 27.46 23.59 24.04 23.79 26.54 23.87 24.81 TiOg.01.02.08.01.02 o006 0.08 Fe203.02.02.14.07.02 - 0.24 FeO.02.02.00.07.01.21 MnO.22.44 2.20.04.04 2.78 2.51 CaO.005.002.004.002.003 none trace MgO.01.015.03.01.01 none 0.05 BaO.0004.0003.0004 0003.0004 n d. n.d. SnO2.019.024.021.026.023 n.d. n.d. SrO.006.006.005.oo6.006 n.d. n.d. Li2O 4.35 5.30 5.24 4.81 4.83 5.05 5.10 Na20.90 1.64 1.10.52.52 0.57 0.52 K20 9.05 9.00 9.35 8,50 9.04 10.14 10.25 Rb20 2.70 2.60 2.50 2.50 2.95 1.62 1.78 Cs20.25.15.11.16.25.09 0.19 H20+ 1.54.62.77 1.00 1.59 1.22 -1.21 H20-.32..6 39.28.31.51 F 6.22 7.79 7.94 6.45 6.69 7.49 6.89 O = F 2.62 3528 3.34 2.72 2.82 3.15 2.90 *Nos. 499, 503, 504, 508, 511, Analyst: R. L. Craig, Electrotechnical Laboratory, U. S. Bureau of Mines, Norris, Tenn.; BaO, Sn02, and SrO determined spectrochemically by CO E. Harvey. No. 7. R. E. Stevens, Amo Mineral., 23, p. 615 (1938). No, 10. A. N. Winchell, Am. Mineral., 27, p. 115 (1942). 165

TABLE XL SPECTROCHEMICAL ANALYSES* OF MICAS FROM NEWRY, MAINE Specimen Nos. 464a 464b 465 483 l Fe203.06.10.23.20 MgO.01.015.01.02 MnO.10.11 o64.09 TiO2 --- - CaO.004.004.37.05 SrO BaO.. — -.0004.0002 Na20.62.33.40.50 Li 20 4.3 4.0 4.1 3.9 RbgO.38.12.52.36 Cs20.08.10.24 o10 Ga203.018.008.007.010 Co203 --- Cr203 SnO2.023.015.001.023 V205 F 6.3 5.8 5.2 6.8 464a: coarse-grained lepidolite 464b: fine-grained lepidolite from same specimenwas 464a. 465: lithian muscovite 483: lepidolite *Analyses by C. E. Harvey 166

TABLE XLI REANALYSIS* OF TWO LEPIDOLITES FROM VARUTRXSK, SWEDEN ANALYZED BY BERGGREN Oxide 521,Specimen Nos. Oxide 521 531 Fe2 0.10.04 253 MgO.008.007 MnO.55.10 TiO2 CaO.002.003 SrO o006.008 BaO.0004 Na20.45.21 Li2O 3.8 4.1 Rb20.48.70 Cs20.09.19 Ga203.013.01 Sc203 23 Cr20SnO2.016.013 V205 F 352 4.8 521 is reanalysis of Berggren (1940) Analysis C. 5531 is reanalysis of No 10 reported in Lundblad (1942): analyzed by Berggren. Spectrochemical analyses by C. E. Harvey 167

TABLE XLII SPECTROCHEMICAL ANALYSES OF LEPIDOLITES FROM TWO OTHER PEGMATITES IN BROWN DERBY DISTRICT Oxide|51 Secien N os.5 - Ox.e514a 14b.. 50o 1 Fe203 ~27.60.09 MgO.008.02.15 MnO 1.2 6.0.13 TiO2.005 o01 - CaO.002.001.032 SrO.006.002.009 BaO.0004 ---.0006 Na20.47.48.72 LiO2 4.4 4.0 3.0 Rb2O.73.46.80 Cs 0.10.10.07 Ga2O3.011.016.03 Sc0 SC2O5 - - --- --- 2 3 Co203. Cr203 - SnO2.005.006.01 V205 F 4.5 5.9 5.8 514a: Untwinned, well-developed portions of lepidolite crystal from Opportunity pegmatite, Gunnison Co., Colorado. 514b: Twinned, poorly developed portions of same book as 514a. 500: Bazooka pegmatite, Gunnison Co., Colorado. 168

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN PART IV. PHLOGOPITE I. MINERALOGY A. Composition 1. General. In phlogopite, K2(Mg,Fe)6(Si6A12)020(OH)4, Na can substitute for K up to nearly K:Na = 1:1 (Harada, 1936 and Heinrich, 1946); also minor Rb,Cs and Ba for K. Fe2 is almost always present but Mg predominates. Fe3,Mn, and Ti usually present in small amounts. The Si: tetrahedral Al ratio varies somewhat and may be larger than 6:2. Heinrich (1946) has shown the relationship between composition and geologic occurrence. Synthetic phlogopite is characterized by F in place of (OH). Manganophyllite may be classed as phlogopite normally with Mg>Mn>Fe and in some types with some Fe3 and Mn3 as well. A complete series exists between phlogopite and biotite. There is no well-defined, natural, compositional boundary between ferroan phlogopite and magnesian biotite. If a division is required, we suggest that where the ratio of Mg:Fe>4:2 the mineral should be classed as phlogopite. Prider (1940) reports a phlogopite from a phlogopite-leucite-lamproite from Western Australia which contains 8.97 percent TiO2. The formula based on the analysis is: (KNa,Ba,Ca)2.06(MgFe2 Fe3 Ti)5.84(SiAl Ti)8.0002o(OHaF)2.16 Prider (1940) notes that the low (OH,F) content is probably not due to loss of volatiles during crystallization at shallow depths and implies that the mineral may have formed from an extremely'dry' magma. He also observes that other comparable micas with high Ti have low hydroxyl content. Tetrahedral positions are occupied by 0.15 atoms of Ti, as Si and Al alone are insufficient to fill these positions; 0.84 atoms of Ti occupy octahedral positions. A discussion of the role of Ti in the mica structure is presented under biotite chemistry. 2. Mahadevite. It has been established that no complete series (solid solution) exists between muscovite and either phlogopite or biotite. In considering these types of micas, Winchell's (1925, 1927, etc.) classification of heptaphyllite and octophyllite is correct. Ramaseshan (1945) and Venkayya (1949), however, have described a mica for which they propose the new species name mahadevite, which they believe to be intermediate 169

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN between muscovite and phlogopite. This conclusion is based upon both chemical and optical data. They note that the high MgO content (13 to 17 percent) eliminates the possibility of the mica being muscovite, whereas the high A1203 content (27 to 30 percent) eliminates phlogopite. The formula for this mica according to Hey (1950) is near (K,Na)o.97(AlFeMg)2.66(siA)4(oo0H)l2. By means of percussion figures the optic plane was determined as being perpendicular to (010). (We have found this procedure unreliable). Through the courtesy of Dr. H. S. Yoder we have been fortunate in obtaining a small amount of mahadevite for X-ray study. From D-, 1- and 2-level a-axis Weissenberg photograph we find that this mica has crystallized as the I-layer monoclinic polymorph, with intensities similar to those of other phlogopites. We also found the optic plane to be parallel with (010), as is natural with 1-layer micas, and 2V to be smaller (<5~) than the value (approximately 14~) indicated by Ramaseshan (1945) and Venkayya (1949). The physical appearance of the specimen is that of a typical phlogopite and there is no interlayered muscovite to account for the high A1203 content reported. Our conclusion is that mahadevite is probably an Al-rich, Mg-, Fe2poor phlogopite, close to Winchell's (1935) end-member eastonite, K2(Mg5Al) (Si5A13 )020(OH) 4. B. Structure 1. Polymorphism. Hendricks and Jefferson (1939) report phlogopite as occuring in 3 polymorphs: the 1- and 2-layer octophyllite monoclinic, and the 3-layer hexagonal polymorphs. This is the only X-ray study of polymorphism in phlogopites reported in the literature. The structures of approximately 80 phlogopite crystals from about 50 different specimens (28 chemically analyzed) were determined by means of the Weissenberg method. All but five have crystallized with the 1-layer monoclinic structure. Of those five, three have crystallized as the 2-layer monoclinic polymorph, whereas the remaining two have crystallized with 3-layer hexagonal structure (Table XLIII). 170

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN TABLE XLIII 2- AND 5-LAYER PHLOGOPITES Number Location Reference Structure 1062 Madagascar Jakob and Parga-Pondal 2-layer monoclinic (a and b) (1932) No. 65 1064 Madagascar Jakob and Parga-Pondal 2-layer monoclinic (a and b) (1932) No. 63 1261 Madagascar Mauguin (1928B) 3-layer hexagonal (b and c) 1220 (a) Hull, Quebec Not analyzed 2-layer monoclinic 1227 (a) Labelle Co., Not analyzed 3-layer hexagonal Quebec It is noteworthy that three of the above specimens are from Madagascar, but that at least four other specimens of Madagascar phlogopite have crystallized with the common 1-layer monoclinic structure. A third crystal obtained from the material supplied by Professor Mauguin (No. 1261 (a)) has crystallized as the l-layer monoclinic form; there probably is little difference in the composition of these several crystals from the same specimen. The specimens of known composition that have crystallized as the 2- and 3-layer forms do not appear to be chemically aberrant. In fact, the compositions of all the Madagascar varieties are practically identical. A few of the phlogopite patterns show diffuse scattering of varying intensity. No correlation of diffuse scattering with polymorphism or chemistry was obtained. These findings are in qualitative agreement with those of Hendricks and Jefferson (1939), who found 14 phlogopites with the 1-layer monoclinic structure and one each with the 2- and 3-layer forms. No new polymorphs of phlogopite have been found in our investigations. Crystals reported by Pieruccini (1950) to be triclinic on the basis of goniometric measurements 171

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN were isolated from a specimen supplied by him. Weissenberg photographs indicate that the structure is the common 1-layer monoclinic form. However, from personal communication with Professor Pieruccini we learn that two distinct varieties of phlogopite are to be found in his specimens. It is possible, therefore, that we did not X-ray the crystals he believes to be triclinic. 2. Relation of Structure to Composition. An attempt was made to correlate chemistry, polymorphism and paragenesis in the dark-colored micas. In all, approximately 60 analyzed biotites, phlogopites, and manganophyllites with varying compositions and different paragenesis were X-rayed. The results, incorporated in Table XLIV, indicates that there is no evident relationship between chemistry, polymorphism and paragenesis among the dark-colored micas. Furthermore, the fact that the three different reported polymorphic modifications found in phlogopites as a whole are found in the Madagascar phlogopites apparently rules out the possibility of temperature or pressure as a major factor; undoubtedly the conditions of formation were fairly uniform for these very similar micas. C. Optical Properties The refractive indices for phlogopite are: a = 1.550 - 1.573 5 -a = 0.028 - 0.049 P = 1.557 - 1.617 2V = 0~ - 12~ (negative) g = 1.558 - 1.618 r<v weak to distinct ca A c = 00 - 40, p = b, /Aa = 0~ - 5~ Optic plane parallel with (010) generally; only in 2-layer forms (which are rare) is it perpendicular to (010). In a few rare types 2V may be somewhat larger than the above indicated range. The indices and birefringences increase with Fe2 and Mn, but change more rapidly for similar increments of Ti and Fe3. Some phlogopites with much Ti or Ti plus Fe+3 have indices as high as a = 1.599,'= 1.645. Phlogopites vary in color from colorless to nearly black or brownishblack in hand specimen. In thin slices many of them are normally colorless, pale yellow, orange, buff or light reddish brown. The depth of the color increases with an increase of Fe2 over Mg and particularly with increasing Fe5 and Ti. The latter probably is responsible for the reddish;tints'that some phlogopites display. 172

TABLE XLIV STRUCTURE OF ANALYZED BIOTITES, PHLOGOPITES, AND MANGANOPHYLLITES ANALYZED BIOTITES Number Reference Structure Paragenesis FeO Fe2O FeMgO TiO2 728 Glass (19355) 3-layer pegmatite 26.72 2.87 n.d. 3.60 729 Stevens (1946) 1-layer pegmatite 8.96 35.531 16.15 1.11 797 Grout (1924) No. 1 1-layer (+3- basic segreg. in 7.72 7.44 16.55 1.67 layer?) granite 798 Grout (1924) Noo 2 3-layer (mod. granite 14.80 4.05 10.21 2.25 scattering) -~ 799 Grout (1924) No. 3 2-layer granite 23.23 3.03 9.24 3.32 -T^ 800 Grout (1924) No. 4 1-layer granite 25.75 1.14 6o16 4.34 801 Grout (1924) No. 5 1-layer peridotite 12.96 8.67 n.d. 1.50 994 vanBiljon (1940) 2-layer granite 13.58 4.23 12.16 3.89 960 Pagliani (1949) 2-layer mica schist 9.10 5.10 10.44 0.56 1001 Inoue (1950) 1-layer nepheline syenite 19.9 4.53 6.26 1.10 1002 Inoue (1950) 1-layer (very nepheline syenite 21.94 8.553 5.32 1.97 heavy scatt.) 1003 Inoue (1950) 1-layer nepheline syenite 21.02 12.45 4.19 lo04 1004 Inoue (1950) 1-layer (mod. cancrinite syenite 16.03 20.22 1.37 0.70 scattering) pegmatite 1010 Kawano (1933) 1-layer metamorphosed xenolith 16.38 3.28 8.99 2.45 1011 Kawano (1942) 1-layer lepidomelane-quartzfels 23.27 7.81 4.32 2.42 1084 Jakob (1931) No. 57 2-layer (weak 2-mica pegmatite 10.47 4.09 13.19 2.06 scattering) 1085 Jakob (1931) No. 58 2-layer 2-mica pegmatite 9.72 2.24 14.22 1.64 1086 Jakob (1931) No. 59 (3-layer?) 2-mica pegmatite 16.26 4.03 8.46 3.16 1087 Jakob (1931) No. 60 2-layer (weak 2-mica pegmatite 16.85 4.08 8.06 2.71 scattering) 1088 Jakob (1931) No. 61 1-layer lamprophre 15.84 5.03 11.17 1.95 1089 Jakob (1937) l-layet feldspar pegmatite 28.61 0.00 6.72 1.99

TABLE XLIV (CONT.) Number Reference Structure Paragenesis FeO Fe203 MgO TiO2 1117 Coats and Fahey (1944) 1-layer (mod. pegmatite (siderophyl- 30.16 tr. 0.22 0.02 scattering lite 1140 C. Oo Hutton (unpubl.) 2-layer from sands derived from 14.49 9.30 5.80 3.47 granitic rocks, Monterey Pen., Calif. 1145 Hallimond (1947) 1-layer marble 6.8 1.1 18.7 1.9 1257 Hutton and Seeyle 2-layer pegmatite-like lense 14.41 3.92 11.11 3.02 (1947) 1262 Mauguin (1928B) 1-layer (heavy from Tschebarkul 12.77 7.09 13.30 1.16 scattering) 1350 Yamada and Sugiura 1-layer (weak pegmatite 3.49 0.67 24.24 0.64 (1950) scattering

TABLE XLIV (CONT.) ANALYZED PHLOGOPITES Number Reference Structure Paragenesis FeO Fe20O MgO TiO2 F * *:* *:.-. - _._ - i~ *: _ I I ~,.: *: - *'' I -, J. /,'':. - ~:, =,'~'~:r _~ i _ ~:. ~ ~!,.~i, -,. ~'~ -- -, 1054 Jakob (1938)? Veiy diffuse from Morawitza 0.64 1.05 26.77 Ool0 0.00 scattering 1055 Jakob (1931) No. 55 1-layer (mod. mica peridotite, 2.52 0.00 25.45 2.80 0.00 scattering) Italy 1056 Jakob (1932)*No. 75 -layer Revision of 55 2.52 0.00 25.45 2.80 0.00 1057 Jakob (1932) No. 70 1-layer Rossie, N. Y. 4.79 0.00 22,30 4.07 0.12 1058 Jakob (1932) No. 69 1-layer Hull, Quebec 2.49 0.71 24.60 1.74 2.04 " 1059 Jakob (1932) No. 68 l-layer Burgess, Ontario 0.30 0000 27.32 0.39 6.74 1060 Jakob (1932) No. 67 1-layer Burgess, Ontario 1.50 0.00 26.14 0.86 2.37 1061 Jakob (1932) No. 66 1-layer Isolo, Madagascar 3.28 000 - 25.29 2.19 1.15 1062 Jakob (1932) No. 65 2-layer (mod. Ampandrandara, 2.96 1.18 23.40 1.69 0.68 scattering Madagascar 1063 Jakob (1932) No. 64 1-layer Mandridano, Madag. 2.09 0.00 24.48 1.64 0.08 1064 Jakob (1932) No. 63 2-layer Saharakara, Madag. 2.79 1.68 23.78 1.11 0.56 1065 Jakob (1932) No. 62 1-layer Ambatoaba, Madag. 1.42 0.66 24.80 0.86 0.86 1066 Jakob (1928) No. 19 1-layer (heavy Simplon-Tunnel 0.00 1.71 24.79 0.39 0.00 scattering 1067 Jakob (1928) No. 20 1-layer dolomite-Tessin 0.00 1.31 25.81 0.83 0.00 1068 Jakob (1928) No. 22 1-layer contact met. car- 2.72 0.97 25.05 0.66 0.58 bonate rock, S. W. Africa 1069 Jakob (1928) No. 23 1-layer Vesuvius 0.58 1.92 28.18 1.27 1071 Jakob-not published 1-layer Skrabbole, Pargas, 0.71 1.01 27.80 0.12 2.12 Finland 1072 Jakob-not published 1-layer Pargas, Finland 5.59 1.54 22.00 1.41 2.80 1073 Jakob-not published l-layer Patteby, Pargas, 1.47 1.29 26.16 0.33 1.87 Finland * Reported in Jakob and Parga-Pondal (1932)

TABLE XLIV (CONT.) Number Reference Structure Paragenesis FeO Fe203 MgO TiO2 F 1074 Jakob-not published l-layer Ontala, Pargas, 1.68 1.87 25.91 0.68 1.30 F inland 1075 Jakob-not published 1-layer Skrabb8le, Pargas, 0.59 1.75 27.22 0.10 0.48 F inland 949 Dana (1892) Anal. 12 1-layer Rossie, N. Y. 7.62 1.12 21.47 1.16 4.00 P. 633 1135 Pagliani (1940) 1-layer in crystalline 1.55 2.65 27,62 2.83 limestone 1139 Hutton and Seeyle 1-layer from marble, New 2.38 0.43 22.95 0.82 0.62 (1947) Zealand 1252 Pieruccini (1950) 1-layer pneumatalytic, 7.89 Tr. 15.66 0.33 2.57 _j~^H~~~~~~~ ~Mt. Somma o\ 1261 f(a) Mauguin (1928B) 1-layer 1 (b and c) Mauguin (1928B)3-layer hexa- Ambotoaba, J 2.30 24.42- 0.74 gonal Madagascar 730 Prider (1940) 1-layer leucite lamproite, 5.75 2.18 19.66 8.97 0.66 West Australia 1325 Cross (1897) 1-layer in Wyomingite 0.90 2.73 22.40 2.09 1.03

TABLE XLIV (CONCL.) ANALYZED MANGANOPHYLLITES Number Reference Structure Paragenesis FeO Fe203 MgO MnO Mn203 TiO2 1076 Jakob (1925, No.8) l-layer contact met. 0.00 16.94 26079 4.52 0.00 0.00 1077 Jakob (1925, No.7) 1-layer contact met. 2054 0.00 29.22 Tr. Tr. 0.00 1078 Jakob (1925, No.6) 1-layer contact met. 0o00 3.95 22.60 0.00 8.30 Tr. 1079 Jakob (1925, No.5) 1-layer Varmland, Sweden 0.00 4.68 21.18 9.25 2,96 0.00 1080 Jakob (1925, No.4) 1-layer contact met. 0.00 2.97 24,80 000 5.77 0.55 1081 Jakob (1925, No.3) 1-layer contact met. 0.00 2.68 26.65 0.00 4.07 0.41 1082 Jakob (1925, No.2) 1-layer contact met. 0.00 0.91 29.28 0.00 3.92 0.00 H 1083 Jakob (1925, No.l) 1-layer Varmland, Sweden 0.00 2.81 27.87 0o00 4.93 0.22 < -

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN D. Inclusions Much phlogopite contains microscopic needles of rutile as inclusions oriented along the principal crystallographic directions. These inclusions are the cause of the well known asterism of phlogopite. They appear to have formed as the result of exsolution similar to much of the oriented hematite in muscovite. The distribution of the inclusions is not always uniform, but may be zonal. Zones with abundant inclusions may alternate with inclusionfree zones. E. Zoning Phlogopites show both color and inclusion zoning commonly. Some examples include the following. Orcel (1926) described a peculiar green phlogopite from Snake Creek, Utah, that displays oscillatory color zoning. A specimen in our mica collection shows alternating light- and dark-green zones. The dominant portion of the irregular crystal is a light yellowish-green with fine darker-green bands parallel with the prism and clinopinacoid. There are minor reeves normal to the color bands. Optical properties are as follows. Orcel This Report np = 1.5529 o = 1.552 P = 1.592 ng = 1.5910 x = 1.593 uniaxial 2V - 2-3 Both zones have essentially the same optical properties. A spectrochemical analysis of a specimen of the light-green material in our collections from the Snake Creek locality is presented in Table XLV. Larsen (1941) reports color-zoned phlogopite of hydrothermal origin associated with aegirine and sodic amphibole in contact-metamorphosed limestone at Iron Hill, Colorado. The micas vary moderately in refractive indices. They are strongly zoned with a pale-brown, or commonly a green core, and one or more outer zones that may be lighter or darker than the core. Some of the outer zones have pleochroism which is the reverse of that normal in phlogopite. Larsen gives the following optical data: 178

TABLE XLV TABLE XLVI SPECTROCHEMICAL ANALYSIS* OF LIGHT-GREEN SPECTROCHEMICAL ANALYSIS* OF ZONED PHLOGOPITE FROM SNAKE CREEK, UTAH PHLOGOPITE, PERTH, ONTARIO Oxide No. 712 1131a - Core 1131b - Rim Si02 40. Si02 41. 40. A1203 12. A1203 12.5 11.5 K20 12. K20 11. 11-5 Fe203 2.3 Fe203 3.6 3.1 MgO 29. MgO 26. 27.5 MnO.15 MnO.06.05 H- TiO2.07 TiO2.85.85 4 GCaO.08 CaO.07.11 SrO -- SrO.006.004 BaO.08 BaO.46.55 Na20 1.05 Na20.75.70 Li20 -- Li20O Rb20 -- Rb20 Cs20 -- Cs20 Ga203.009 Ga203.004.004 Sc2O3 — Sc203 Co203 oo004 C0203.0008.0009 Cr203.015 Cr203.0027,0026 Sn02 - S 02 V205 - V205.017.0o14 F.535 F 2.7 3.7 * Spectrochemical analysis by * Spectrochemical analysis by C. E. Harvey C. E. Harvey

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Specimen IH - 2 core: a = 1.576 P = 1.613 1 = 1.613 pleochroism a nearly colorless, P and' faint brown intermediate zone: a = 1.576 p = 1.613 k = 1.614 pleochroism a deep chestnut brown, P and K nearly colorless outer zone: pleochroism a deep chestnut brown, P and' nearly colorless Generally, the indices of refraction are the same for the different zones, but the axial angle may be a little larger in the outer zones. A phlogopite (1369) from Thorne Township, Quebec, Canada, shows four concentric color zones. The specimen varies in color from a deep red-brown at the core to a light green-brown at the margin. The boundaries between zones become sharper toward the core. Optical data for each zone are as follows. Zone 0 P_ J 2V Color 1 (core) 1.575 1.622 1.622 1~- 2~ deep red-brown with many rut ile inclusions 2 1.575 1.622 1.622 1~- 2~ deep red-brown with no inc lusions 3 1.568 1.611 1.611 0~ red-brown, inclusion free 4 (margin) 1.559 1.599 1.599 0~ green-brown From margin to core, each zone shows a progressive increase in indices and size of the optic angle. Another phlogopite (1370), an irregularly shaped sheet from an unknown Canadian locality has a parallel arrangement of light-brown inclusionfree zones alternating with darker-brown zones containing numerous inclusions of rutile. Both zones are essentially uniaxial. The inclusion-free zones show no asterism and yield a sharp figure, whereas the zones bearing inclusions show strong asterism and yield an extremely diffuse figure. Indices of refraction for each zone are: Clear zones 1.541 1.570 Zones bearing inclusions 1.544 1.578 180

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Two analyses were made of a zoned, amber-colored phlogopite (1131) from Perth, Ontario, one from the core of the sheet, the other from the rim (Table XLVI). There is a slight difference in Fe, none in Ti, a distinct difference in F. Numerous specimens of Canadian phlogopite available to us display color zones. Generally, the optical properties of the different zones in a single crystal are essentially the same. Wherever a variation was noted, the darker-colored zones or zones of inclusion have higher indices of refraction and a larger optic angle. More specimens of phlogopite tend to be darker toward the center and lighter toward the rim than the converse. 181

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN II. OCCURRENCE A. Nonpegmat itic Phlogopite or phlogopitic biotite occurs in ultramafic igneous rocks, particularly kimberlites. It also forms as a secondary constituent in many other types of peridotites, probably largely through autometasomatism or hydrothermal metamorphism. Phlogopite also is a primary constituent of certain lamprophyres, mafic and ultramafic potassic rocks, and a few rare extrusive types. In metamorphic rocks phlogopite is formed in marbles or Mg-silicate gneisses by the regional metamorphism or contact metamorphism of argillaceous dolomites, or impure magnesian limestones. If the phlogopite contains abundant F, metasomatism is generally indicated as a means of introducing that element. However, not all phlogopites contain fluorine and the mica may form without additions to the system, if sufficient water is available. Phlogopite marbles are considered to belong to the amphibolite facies. Phlogopites occur also in various skarns in whose formation metasomatism has been important. B. Pegmatitic and Vein Occurrences Phlogopite deposits occur as veins consisting chiefly of phlogopite, apatite and calcite in pegmatite and pyroxenite in southern Quebec and eastern Ontario. Scapolite, actinolite, diopside and many rarer minerals occur as accessory constituents. The veins and concentrations occur in or adjacent to masses of pyroxenite that have been intruded by bodies of granitic, microcline-rich pegmatite. Landes (1938) believes that the deposits were formed from the hydrothermal phases of granitic pegmatites markedly contaminated at depth by dolomitic linestone. Analagous deposits occur in Madagascar in pyroxenite but here apatite is not abundant (Savornin, 1934). Other vein deposits of phlogopite, calcite, apatite, pyroxenite, and diopside transect impure marbles, amphibole schists, and amphibolites in the Sejudjanka region of the USSR where these rocks have been intruded by pegmatites and granite dikes (Ginsburg, 1930). 182

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN III. MANGANOPHYLLITE A. Introduction Manganophyllite was first described by Igelstrom (1872), his specimen having come from the iron and magagnese mines at Pajsberg near Filipstadt in Sweden: "The manganophyllite is accordingly the manganese-rich mica. The one known up to now is placed near phlogopite or biotite. Perhaps one can take it for a manganese-rich biotite." Dana (1892) states: "In ccmposition manganophyllite is a tuanganesian biotite, but varying widely in the relative amounts of manganese and other elements." B. Structure The only X-ray work reported in the literature on manganophyllite has been that of Hendricks and Jefferson (1939). They report a manganophyllite from Pajsberg, Sweden with a 2-layer monoclinic structure. We have found that all the Swedish manganophyllites described by Jakob (1925A)have the 1-layer monoclinic structure. A specimen from Franklin, N. J. (No. 1358) has a 2-layer structure, however. The Swedish manganophyllites that have a very small 2V (particularly Jakob, 1925A)No. 1 and 5) appear to approach the 3-layer hexagonal structure owing to an intimate "twinning" of thin sheets similar to that described for uniaxial lepidolites. The chemical analyses of these micas show no correlation between composition and polymorphism. Manganophyllites with the 1-layer form show extreme ranges in- the Fe and Mn (for example, Fe203 = 0.00 to 16.94 percent; see also Table XLIV). Jakob (192%A)reports most of the manganese in these micas as Mn203. C. Composition The general formula for manganophyllite can be written as follows: K2 (Mg5-4,Mn2 - lFe3o o.5, Feo30.Mn30-1) (Si6Al2 )020(OH)4 with a total of 6 atoms in octahedral positions. Other possible iscmorphous substitutions are minor NE, Ca,,and Ba for K; and minor Ti for Si in tetrahedral positions or more probably in octahedral positions for MgMn, or Fe3. In biotites small amounts of Mn are associated either with high ferric or ferrous iron. Some manganophyllites contain both Mn2 and Mn3, 183

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN others have only Mn2, and in a few the manganese is chiefly divalent. Manganophyllites are low in iron and have a maximum of about 18 percent MnO plus Mn20O. Phlogopites on the other hand have about a maximum of 50 percent MgO and less than 10 percent Fe and Fe203 combined. Some investigators class manganophyllite as a variety of biotite, but owing to the small amounts of Fe2 and large quantities of Mg, most manganophyllite should be classed as manganian phlogopite. D. Optic s a = 1.548 - 1.575 3- a = 0.024 - 0.040 P = 1.581 - 1.636 2V = 0~ - 34~ Y = 1.581 - 1.636 There appears to be no generalization which accurately represents the relationship between indices and other optical properties with composition. For example, Jakob's (1925A)analysis No. 8 with 16.94 percent Fe203 has the highest indices of any of his manganophyllites. However, Jakob's analysis No. 6 has indices approximately 0.03 lower and contains only 3.95 percent Fe203. The Y indices of a few manganophyllites were plotted against 1/2 (Mn203 or MnO) + 2(Fe203 + TiO2) by Heinrich (1946). The results indicated that the effect of the manganese on the index of refraction is less than that of the ferrous iron. The most significant observation concerning the optical properties of the manganophyllites studied is that several of Jakob's (1925A)specimens (Numbers 3, 4, and 6) showed the optic plane perpendicular to (010). Practically all micas crystallizing as the 1-layer polymorph have the optic plane parallel with (010). The orientations of these micas were determined by X-ray methods; the cause of the anomalies has not been determined. E. Single Crystal Variation A survey of the literature failed to uncover any information regarding structural or optical variations or color zoning in single crystals of manganophyllites. In our work we have found no cases of polymorphic variations or color zoning within single crystals of manganophyllites. Small-scale variations in the size of 2V within individual crystals have been observed, but are generally insignificant. L__________________________________ 184

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN F. Occurrence In Table XLVII are presented a listing of all the occurrences of manganophyllite (including doubtful or incompletely described manganophyllites) we have been able to find in the literature. G. Conclusions 1. Manganophyllite is a magnesium-manganese mica, with variable amounts of ferric. and ferrous iron crystallizing, with 1-layer and 2-layer monoclinic structures. 2. Most manganophyllites could be described equally well as manganian phlogopites. 3. Many reported manganophyllites are of doubtful authenticity and need further study, particularly chemical analyses. 4. Most manganophyllites of undoubted authenticity occur in manganese deposits of contact metamorphic origin. 185

TABLE XLVII OCCUREENCE OF MANGANOPHYLLITE Assoc. Location Type Deposit Host Rock A Other Minerals Reference Remarks Mn Minerals Ada Co., disseminated pegmatite - quartz, Kauffman, et al. Probably a Idaho orthoclase (1950) biotite; spectroanalysis given Mason Co, lenses, pock- volcanics, bemenite, neo- calcite, jasper Park (1942) Washington ets, and thin limestones, tocite and other Pardee, et al. (1921) beds and argillites Mn silicates and minerals Kaso Mine, veins of basic quartzite and rhddonite, teph- Mn-bearing biotite, Yoshimura (1939) Called mangan H Japan ccmposition slate roite actinolite and phlogopite; 01 other silicates chemical analand Ba-feldspar ysis given Franklin, disseminated skarn many Mn-bearing Zn and Fe Chester (1894) N. J. silicates minerals Palache (1935) Langban, disseminated skarn many Mn-bearing Fe ores Palache (1929) Sweden silicates Jakob (1925A) Earstigen, For chem. anal. see: Occurrence Pajsberg and Jakob (1925A) similar to Jacobsberg, Dana (1892) that of Sweden Langban Ankara, Turkey sedimentary rhodonite, rho- Wijkerslooth (1943) Need further rocks dochrosite, study braunite Central disseminated mica schists in Mn deposits Fermor (1909) Incampletely India described Postmasburg, sedimentary psilomelane, Hall (1926) Need further South Africa rocks pyrolusite, Chudaba (1929> study rhodonite _~ * - -. * _J7 * *, _, j, _ _ ~, _ L _.... L.........'- - - JL

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN PART Vo BIOTITE I. MINERALOGY A. Composition K2(Fe2Mg)64 (Fe3,Al,Ti)02 (Si6-5Al23 )0-22 (OH,F)4-2; some Na,Ca,Ba,Rb, and Cs for K; Mn for Fe2. Mg may be almost absent in some varieties; others contain much Fe5; in some Li may replace Al possibly leading to tinnwaldite. Zambonini (1919) reports 14.33 percent CaO (see p. 42 ). Freudenberg (1919) reports a biotite with 12.5 percent TiO2. It has been pointed out that a complete series exists between biotite and phlogopite; a suggested division is discussed under phlogopite chemistry. For a discussion of the role of titanium in biotite (and micas) and the variation of tetrahedrally coordinated Al/Si, the reader is referred to the section on the relation of composition and occurrence (p. 195 ). B. Structure 1. Introduction. Hendricks and Jefferson (1939) report biotite as occurring in five polymorphs: 1-layer monoclinic - space group Cm a = 5.35 b = 9.2i co = 10.2K P = 100~ 2-layer monoclinic (octophyllite type) space group C2/C a = 553A b = 9.02 co = 20.2A 5 = 95~ 3-layer rhombohedral enantiomorphic hemihedral - space group C3112 or C3212 a = 5.3A co = 30.0A P = 90~ 6-layer triclinic holohedral - space group PI a = 5.3A b = 5.3A c = 60.0A a 9o~ ~ = 9o~ =90 120~ 187

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 24-layer triclinic holohedral - space group P1 a = 5.3A b = 5.53 c = 240.0o a = 90o~ = 90~ = 120~ The only other polymorph reported in the literature is an incompletely described 18-layer triclinic biotite (Amelinckx, 1952B). 2. X-ray Technique. The X-ray study of biotites, and to a lesser extent of phologopites, is more difficult than that of muscovites and lepidolites. A crystal orientation procedure based on optical methods was perfected and used satisfactorily with the latter series. The positions of the crystallographic axes were determined by means of interference figures and extinction positions, for in muscovite and lepidolite the crystallographic axes correspond essentially to extinction positions. In phlogopite and biotite, however, the birefringence of (001) sections in often nil and 2V approaches zero; thus an optical orientation method is impractical. To overcome these problems, Laue photographs were used to determine the positions of the desired axes. However, Laue photographs of some monoclinic specimens appear so nearly hexagonal that the true plane of symmetry is not easily determined (owing to the large crystal-to-film distance of our camera, few reflections were recorded). Weissenberg photographs about the two pseudo a-axes were often taken before the one about the true a-axis was obtained. An alternative method was used occasionally, whereby it was not necessary to take a Laue photograph. This consisted of producing a precussion figure on the cleavage flakes and choosing one of three rays at random as the rotation axis, with the hope that it would be the true a-axis and not one of the pseudo a-axes. But, as in the first method, several attempts usually were required before the structure could be determined accurately. Zero-level Weissenberg photographs taken about one of the pseudo aaxes of the 2-layer monoclinic polymorph appear to be identical with the 0level a-axis photograph of the 6-layer triclinic biotite illustrated by Hendricks and Jefferson (1939, p. 745). Likewise, 0-level pseudo a-axis photographs of the 1-layer monoclinic polymorph strongly resemble 0-level b-axis photographs of the 6-layer monoclinic polymorph reported only in lepidolites. Therefore extreme caution had to be practiced in deciding the structure of these micas from O-level photographs. Diffuse scattering, particularly among the biotites, also made structure determinations difficult in several instances. In some specimens the crystal development of the flakes was poor, or the only analyzed specimens available were powders. Under these conditions, single-crystal methods were not possible, and the powder X-ray method was 188

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN attempted. It was found that iron radiation was most suitable and even essential in studying biotites by the powder method. A filter was not used, for it increased the necessary exposure time by a factor of nearly three. Unfortunately the powder patterns of the common polymorphic forms are so similar that no distinctive patterns could be obtained. A monochromator, for longer exposures without background fogging, was not available. 35 New Structural Data. Although the structures of about 200 biotites have been determined, only a few definite conclusions are available. Attempt-s:-t-J:'rel;' te'-Cehemistry withwa$:uB.a-er,-ave beenQ. ~. yl gfsul, but( ) ),.i information is available on the geographic and geologic(di.ottribAlmon of various polymopPhs. The reader is referred to Table XLIV in which the results of the X-ray study of 27ayzed biotites fid;es = apparent rebls A hi ip be-(- ) -i tween c-mi:, anGpyrOmorphism in tejbbitites (and ghlogopites).'l:.-'aIn.r.e t,.ai.cd-of the biot,.tes t-xijtfctures of se e~.l~common( x:' varieties were determined, The results are tabulated in Table XLVIII.. During -thjis study neither the 6-layer nor the 24-layer triclinic biotites described(by iendricks an Teff~eb(i959) were fouyi a-end-r stud!\ifi' of thes LrqEui t -Plrlayer biot t i S bscIj-so Hendricks and. Jefferson,9). from Sterling, New York (U.S.N.M. C367) indicates the l-layer monoclinic struclittit',: D3:ot' -S._JL:-Ybder (pe.~rcz. ti, ation) h,6pSt: ter mape l. from the same specimen and has found the 2-layer monoclinic structure. Only the 1 cid' -l'iy -'Aclinic,~-'-e... ae~-1-.xagonaic:4.~... prT::were — r nitely es..si;. Id^' In a few scattered instances patterns not assignable to tnitely e ir' i9J3BI any known polymorphs were recorded on Weissenberg photographs. Detailed study of thet er n s.areriNea,11 probably io t that they, consist Of ystals with mit es structures rather than new polymorphs.. crystallized. as thle layer monocclinicpI h 15 as thI,:t'h hex-()t4r1 agiO3 py;m~gT1 btut~. nly 1 as the 1-l> moboclinic p olyroi h m Heavy daift. fuse scattering prevented accurate structural determination o ue re8iaioner&Diffusescattering is characteristic of biotites from this area. The results of studies from other widely scattered pegmatite districts aret asum^inarze below: Southern Norway - predominantly 1-layer monoclinic forms, but 2-layer forms abundant. Canada: Iancroft, and Wilberforce Districts - predominantly 1-layer forms. Colorado: Guffey Districta;fpredominately 1-layer forms. 189

TABLE XLVIII STRUCTURE OF BIOTITE VARIETIES Number Name Locality Structure 732(a) waddoite Isle of Waddo 1-layer monoclinic 746(a) lepidomelane Brevig, Norway 2-layer monoclinic (pterolite) 752(a) meroxene Mt. Vesuvius 1-layer monoclinic 753(a) meroxene (white) Mt. Vesuvius 3-layer hexagonal 759(a) meroxene Mendham, N. J. 1-layer monoclinic 754(a)' calciobiotite Italy 1-layer monoclinic 770(a) annite Rockport, Mass. 1-layer monoclinic 1116(a) annite Rockport, Mass. 1-layer monoclinic 775(a) siderophyllite Pikes Peak, Coloo 1-layer monoclinic 1117(a) siderophyllite Brooks Mt., Alaska 1-layer monoclinic (very diffuse scattering) 783(a) cryophyllite Rockport, Mass. 1-layer monoclinic 1112(a) monrepite Finland 1-layer monoclinic 1113(a) eukamptite Presburg, Hungary 1-layer monoclinic 1114(a) bastonite Bastogne, Belgium 2-layer monoclinic 1130(a) haughtonite Scotland (?) - mixed structures 190

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Geological environment may play an important role in the crystallization of biotite micas. The pegmatites of the southeastern United States (North Carolina, particularly) are of quartz diortic composition, whereas those of the Bancroft area are nepheline syenitic. It seems possible that biotites crystallizing from such different magmas might have structures characteristic of their particular environments, but this hypothesis needs considerable further study. As mentioned above, composition alone does notseem to be the exclusive factor. It was hoped that a knowledge of the temperature of crystallization of these micas would be very useful, and for this reason a suite of biotite phenocrysts from hypabyssal acid dike rocks from the Guffey District, Colorado, was studied. However, both 1- and 2-layer polymorphs were found in these biotites. C. optical Properties 1. Indices, 2V, Optic Plane; Variation with Composition. The refraction indices of biotite are: a = 1.565 - 1.625 Y-a = 0.040 - 0.060 = 1.605 - 1.675 2V = 00~ 25 g = 1.605 - 1.675 (negative) 2V is usually less than 10~, but a few rare varieties have 2V somewhat greater than 25~. Dispersion: r> v or r < v a Ac = 8~- -2~, = bj Y\ a = 0 -9~ The optic plane is usually parallel to (010)) less commonly (only in 2-layer polymorphs) normal to (010). The 1-layer polymorph is the most common structure and invariably has its optic plane parallel to (010). The indices increase with Fe2 and more rapidly with Fe3 and Tio For example, see the studies of Winchell and Winchell (1951, p. 3749 Fig. 257) and that of Heinrich (1946), in which it was demonstrated that the effect of Fe3 and Ti in increasing the indices of refraction is approximately twice that of Fe2 and that the effect of Mn3 was approximately one-half that of Fe2. In biotites in which total iron remains constant, 1 percent Ti02 increases refractive indices 0.0046 (Hall 1941). In some rare ferrian biotites r may reach as high as 1.73 and'-ca = 0.080, 2. Color; Variation with Composition. Biotites are deeply colored, usually in shades of brown and green. Red-brown biotites owe their color mainly to combined titanium and ferric iron; greenish-brown 191

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN - and brown biotites are colored mainly by Fe2 with lower amounts of Ti and Fe3; and blue-green biotites are colored mainly by Fe2 (Hall, 1941 and Shibata, 1952A). 3. Zoningo Zoning in biotite has not been as widely recognized as in muscoviteo This may be due to the much darker color of most biotites in which slight color differences would be hidden except in very thin sheets. Some examples of zoned biotites reported in the literature are listed. Brogger (1921) reports optically zoned biotite from the Fen Region of southern Norway, in damkjernite, a porphyritic rock containing phenocrysts of biotite and pyroxene in a fine-grained groundmass of pyroxene, biotite, and magnetite with minor nepheline, microcline, and calcite. In hand specimen the biotite isbronze-brown, but under the microscope it is reddish in color. The biotite flakes are not six-sided, but usually irregular. Pleochroism of the core is from ruby-red or orange-red to bright yellow; pleochroism of the rim is from bright green to bright yellow or colorless. The biotite seems to be uniaxial. Johannsen (1937) notes the occurrence of color-zoned biotite in minette, a lamprophyre of the syenite series, composed of bibtite phenocrysts in an orthoclase groundmass. The biotite is dark brown under the microscope and in both the phenocrysts and the groundmass it shows "a light center surrounded by an iron-rich opaque border." Grout (1924) mentions that biotite in the Minnesota granites commonly shows a zonal structure. Clarke (1887) describes a bronzy-black mica (siderophyllite), from Pikes Peak, Colorado, "...resembling phlogopite externally,... but made up of a core composed of a soft rotten material, evidentally derived from the original mica, and surrounded by a broad black margin of the latter. Streaks of rusty alteration products reached into the margin in every direction." Analyses for both zones are given as follows. Margin Center H20 4.54 7.28 SiO2 34.21 34.63 A1203 16.53 17.95 Fe203 20.12 31.25 FeO 14.17 3.01 MnlO 0.91 0.34 CaO 0.48 0.81 MgO 1.34 1.08 Na20 1.43 0.89 K20 6.50 1.96 F 0.08 0.54 192

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Coats and Fahey (1944) describe an optically zoned siderophyllite from an alaskite pegmatite near Brooks Mountain, Alaska. In thin section some of the plates show a core having a double refraction of 0.050 and 2V = 6~ - 8~ surrounded by a discontinuous rim of pale-blue mica having a double refraction of 0.028 and 2V = 37~. Over 100 specimens of biotite were examined for zoning. None of the pegmatitic biotite showed any zoning. One thin section of minette from Freiberg, Saxony, contained phenocrysts of zoned biotite as described by Johannsen (1937). Apparently zoning in biotite is very uncommon. 195

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN II, OCCURRENCE A. Igneous Rocks Biotite occurs as an essential or accessory mineral of most types of intrusive igneous rocks, particularly pegmatites, aplites, granites, granodiorites, syenites, feldspathoidal syenites, monzonites, diorites and norites; less commonly in diabases, other gabbroic types, and in rarer types. In many of these it appears as a primary constituent, although in some, such as diorites and gabbros for example, it has apparently formed by deuteric reactions and can be seen as replacements of and overgrowths on such minerals as hornblende, augite or magnetite. Biotite, as phenocrysts, also is widespread in felsic and intermediate extrusive rocks, and may show strong resorption mainly to iron oxide and pyroxene. Biotite is rare, however, as a matrix mineral in these rocks. B. Metamorphi c Rocks Biotite is a widespread mineral of metamorphic rocks and is stable under a broad temperature-pressure range, forming apparently at relatively low temperatures and remaining stable into the granulite facies. In contact metamorphism it forms under low grade conditions in metamorphosed argillaceous, impure arenaceous, as well as in intermediate and acid igneous rocks and remains stable into the higher intensity environments. The biotite zone is indicative of moderate-grade regional metamorphism in argillaceous rocks, but, if stress is abnormally low, biotite may form even at somewhat lower temperatures. The reaction between chlorite and muscovite yields biotite for such rocks as biotite schists, biotitesericite schists, biotite-chlorite schists and albite-biotite schists. Once formed, biotite usually persists into higher-grade zones, appearing with garnet, kyanite, sillimanite, and cordierite, in schists,; gneisses and granulites. More mafic igneous rocks and somewhat calcareous sediments also yield biotite-bearing types upon regional metamorphism - biotite-plagioclase schists, biotite-hornblende schists, and biotite-epidote schists, for example. C. Variation of Composition with Occurrence Heinrich (1946) has shown that biotites of intrusive igneous rocks vary systematically in composition with the rock in which they were formed. This variation is chiefly an increase in the Fe2/Mg ratio toward the more felsic rocks. Biotites rich in Fe3 are strongly represented in alkalic rocks, 194

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN and also in extrusive igneous rocks where some of their Fe2 was oxidized to Fe3 by escaping gases. Others have traced the successive changes in individual elements in biotites. Chapman and Williams (1935), for example, in their study of the alkaline White Mountain magma series, have observed that in going from a gabbro to granite, the biotite will show a decrease in 8i,Ti,Al, and particularly Mg with a corresponding substantial increase in Fe. Similar results are reported by Inoue (1950) on biotites from nepheline syenites and nepheline syenite pegmatite. She also notes that biotites from nepheline syenites tend to be higher in Mn. Nockolds and Mitchell (1948) in a comprehensive study of the non-alkaline Garabal Hill - Glenn Fyne complex show a decrease in SiTi, and Mg with an increase in Al,Fe3, and Fe2 in the biotites on progressing from pyroxene-mica diorite to porphyritic granodiorite. In their study of the trace elements they find a progressive increase in GaMn, and Rb on passing from early to late biotite whereas Cr,VCu,Ni,Co, and Sr decrease in the same direction. Ba remains more or less constant in early biotites and decreases in the late stages; Li, on the other hand, increases rapidly in the early biotites and later remains more or less constant. The role of Ti in the crystal lattice of the micas is discussed by Serdiuchenko (1948B). He concludes that at low temperatures and in an acid medium, Ti proxies for Mg, whereas at high temperatures and with an increase in the alkalinity of the medium, Ti may substitute for Si. Thus, in alkaline igneous rocks Ti may be found in the tetrahedral layer or may enter both the tetrahedral and octahedral layers. In nonalkaline igneous rocks Ti is exclusively in the octahedral layers and isomorphously replaces Mg. Biotites of feldspathic quartz-free rocks (mainly lamprophyres) occupy intermediate positions with relation to Ti (but closer to the position of Ti in alkaline rocks). The presence and role of Ti3 in phlogopite and biotite is recorded and discussed by Jakob and Parga-Pondal (1932) and Jakob (1937). However, they maintain that Ti is not found in the tetrahedral positions. Most workers believe that titanium replaces magnesium. Several analyses, however, are recorded in the literature in which some of the Ti must be grouped with silicon and tetrahedral aluminum in calculating the formulas so that all tetrahedral positions are filled. Such a case is well illustrated in the study of a titaniferous phlogopite described by Prider (1940); Serdiuchenko (1948B) gives about 10 other examples of biotites and titanian biotites (wodanites). Serdiuchenko (1948B) also finds it necessary to include small amounts of Fe3 to fill tetrahedral positions in the case of three lepidomelanes. In this manner the atomic requirements for the tetrahedral layer are completed with Ti (and Fe3), since the total of aluminum and silicon is insufficient. Nockolds and Mitchell (1948) also discuss this problem and point out that practically all the aluminum in the biolttes they studied (from diorites, granodiorites and granites) is tied up in tetrahedral positions. 195

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Serdiuchenko (1948B) (in translation) quotes Fersman as reporting: "Ti replaces Si in tetrahedra of the pyroxenes, amphiboles (chains) and micas (sheets) but not in feldspars or zeolites (5 dimensional linkage). However, the crystallochemical role of titanium in the lattice of minerals is not determined only by the type of structure, but for the very same minerals (for example micas, amphiboles and others) its role changes with respect to the thermodynamic and geochemical conditions of the genesis (temperature, pressure, oxidationpotential' pH)." Machatschki (1930) states that at low temperatures Ti may not replace Si because of the differences in ionic radii of titanium and silicon atoms. At high temperatures, however, it may replace Si, but the tendency for exsolution of ilmenite or other titanium minerals appears to be great. Bragg (1937) points out that Ti is exclusively found in sixfold coordination and therefore it is very unlikely that in the micas it would replace silicon in fourfold coordination. Ramberg (1948) believes that Ti is relatively abundant in high-temperature metamorphic biotites and accounts for their stability. In our opinion the problem still remains unsolved. The accuracy of the conclusions reached by Serdiuchenko (1948B) depends upon (1) whether it is absolutely necessary for the tetrahedral positions to be filled (the octahedral may be deficient), and (2) the accuracy of the analyses used in the compilation:. In metamorphic rocks the data are still too few to permit much correlation between composition and the temperature-pressure environment. Undoubtedly, howeer, metamorphic biotites vary considerably in composition. It has been found in a review of the literature that biotites formed under conditions of simple thermal metamrphism generally have a n.Ah larger quantity of Al in tetrahedral position than biotites formed under t dition of regional metamorphism (Dr. A. Miyashiro, personal communidation). This fact appears to be in accord with the general tendency of minerals that formed at high temperatures to contain more fourfold Al than those that formed at low temperatures. Harry (1950) has recently demonstrated that this is the case for some igneous and metamorphic minerals, particularly amphiboles and pyroxenes, and believes that the situation is analogous in biotites. He suggests that large aounats of tetrahedrally coordinated Al may help stabilize biotite that fored under physical conditions (high temperature) ordinarily promoting the 196 -

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN crystallization of simpler lattice-type minerals. This suggestion is based upon the observation that biotite has the largest amount of Al substitution for Si in a rock also containing pyroxene and hornblende in which all three minerals crystallized under the same pressure-temperature conditions. ----------------- 197

- ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN III. PEGMATITIC BIOTITE A. Localnzation The occurrences of pegmatitic biotite may be grouped under 1. Wall-zone occurrences 2. Outer intermediate-zone occurrences 3. Fracture-fillings and fracture-controlled replacements 4. Exomorphic and endomorphic biotite. Biotite of magmatic origin occurs normally in the outer zones of pegmatites. In wall zones the biotite may occur in elongate blades or plates that tend to be arranged normal to the wall-rock contacts. Not uncommonly these blades are "tangle foot", i.e., they contain a central strip of biotite and bordering parallel overgrowths of muscovite. The School Section pegmatite of Eight Mile Park, Colorado, contains such blades abundantly in the hangingwall section of the wall zone. Many pegmatites of the southeastern districts contain unoriented books and plates of biotites in wall zones. Smaller books, flakes and scales of biotite also are abundant in outer intermediate zones in many districts of New England and other regions. In such zones the biotite occurs with quartz,. microcline, sodic plagioclase and, in some cases muscovite as well. Fracture-fillings of biotite are conspicuous features in some pegmatites, for example, those of Topsham, Maine; in the Big Flint pegmatite, Jackson County, North Carolina; in the Yard pegmatite, near Salida, Colorado, and in many pegmatites of other districts, especially some in Southern Norway and Sweden. The biotite crystals are commonly of large size, 1 to 20 feet, and usually appear in subparallel to slightly rrdial patterns. Dendritic arrangements also occur with a central large sheet as a stalk and smaller, sheets as branches. In some districts with biotite of this type, rare-earth minerals, such as fergusonite, gadolinite, samarskite, etc., occur relatively abundantly, as crystals and grains between the biotite sheets. In a few deposits small books of muscovite have been formed between the larger parallel biotite sheets, normal to them. The outer parts of some pegmatites contain wisps and schlieren of recrystallized coarse-grained parallel biotite derived by incorporation of blocks of biotite-rich wall rocks. Examples are the Ledford Cove pegmatite, 198

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Franklin, North Carolina and the Big Ridge pegmatite near Waynesville, North Carolina. Most pegmatites that have been intruded into hornblende-rich rocks have developed prominent selvagesof coarse biotite by reaction. In a few pegmatites more than one generation of biotite has been observed, e.g., wall-zone plus fracture-filling biotites, but there is no information as to any compositional variation between the two types. It has been observed that pegmatites containing concentrations of green book muscovites in wall zones rarely contain any biotite, except some that may be exomorphic in origin. Biotite associated with or included in green book muscovite usuall itself is deep olive green in color, whereas that intergrown or occurring with ruby muscovites is black or dark brown, probably with more Fe3 and Ti and less Fe2. Where abundant biotite occurs with muscovite in zones, the muscovite normally is largely free of iron-oxide inclusions. In some districts biotite and schorl are mutually exclusive, either in wall zones or in the entire pegmatite. B. Macrostructural Features A systematic study of the structural features of book or sheet biotite has not been attempted. Biotite commonly forms in thin sheets or flakes from a fraction of an inch to many feet across. Thick books are relatively uncommon. Ruling marks thicker books in some deposits. Many sheets that crystallized along cracks follow the curvature of the fractures. Some very large sheets of this type show' a diamond-shaped crosshatch pattern of markings resembling reeves but less closely spaced.'Intergrowths with muscovite are common, normally with the biotite occupying the central part of the composite book or sheet. However, an overgrowth of central muscovite with framing biotite also is known. Concentric structure in biotite has been reported by Simpson (1912) from a pegmatite in Ubini, Western Australia. 199

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN PART VI. ZINNWALDITE I. MINERALOGY A. Composition General composition: The composition of zinnwaldite may be expressed by the following general formula: K2(Fe21-2,Li2-3JA12) (Si6-7,A12-1)0 20(F3- 2, OH1-2) - The (AlFe,Li) group usually contains six atoms but may be considerably deficient. The possible ionic substitutions that can take place are as follows: For Si: One-fourth of the Si may be replaced by Al; minor Ti may also be substituted in this position. For Al: TiFe2,Fe3,MnMg, and Li. For K: Na,BaRb,Cs, and minor Ca. Trace elements: Glass (1935) in studying zinnwaldite from Amelia, Virginia, reports the largest variety of trace elements ever recorded in a zinnwaldite: B,Be,Sn,Zn,Rb,Li,Cs,Ti Cs20 = 0.10 percent and Rb20 = 1.04 percent. Other trace elements in zinnwaldites are P,He,Mg,Mn3,Ga,Ba,Sc,T4 and Y (Rankama and Sahama, 1950). The chemical similarity between zinnwaldite, lepidolite, and biotite are presented in Table XLIX. TABLE XLIX CHEMICAL RELATION OF ZINNWALDITE TO LEPIDOLITE AND BIOTITE -%.Typialical -i a -i Z ~i nn~aia Leo-lite ir__.. m- L, P.,.'_1'.' _;-doiif -,; fB 1 o t i,e SiO2 48 50 35 Al203 20 23 20 FeO 8 - 8 Fe20 2 -- J Li20 4 5 - F 7 7 -- 200

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Zinnwaldite is, in general, chemically closer to lepidolite than biotite, with the important exception of Fe. B. Structure Three structural varieties of zinnwaldites are now known: (a) 1-layer monoclinic, (b) 2-layer monoclinic muscovite-type, and (c) 3-layer hexagonal. Hendricks and Jefferson (1939) report zinnwaldites with the 1-layer structure from Amelia, Virginia; Zinnwald, Bohemia; and Brambach, Saxony. They also report a 3-layer hexagonal structure zinnwaldite from Amelia, Virginia. The X-ray data of the zinnwaldites studied by us is tabulated in Table L. Eight samples studied have crystallized as the 1-layer polymorph, five as the 3-layer polymorph, and one as a 2-layer polymorph very similar to that of normal muscovite. An excellent Weissenberg photograph obtained from a flake of zinnwaldite (Spec. 1251) from UIbini, West Australia, (Murry and Chapman, 1931) indicates the 2-layer monoclinic structure with the presence of (06) reflections with I odd (Fig. 15). The significance of these reflections has been discussed in detail by Hendricks and Jefferson (19539),and in this report with regard to lithian muscovite. The significant point is that (06i) reflections with Yodd are present in the 2-layer muscovite (heptaphyllite) type of micas, and their presence indicates considerable distortion from the ideal mica structure. Hendricks and Jefferson (1939) believe that the muscovite-like micas produce these reflections because of an incomplete filling of the octahedral positions which results in the distortion. Muscovite has only two-thirds of the octahedral positions filled, whereas the ideal octaphyllite micas have all these positions occupied. Therefore, the presence of 06J reflections with I odd recorded on the Weissenberg photograph of Spec. 1251 implies that there is distortion in the structure, and that this zinnwaldite approaches the heptaphyllites in structure This appears to be the only specimen of zinnwaldite which has been definitely shown to be more closely related structurally to heptaphyllite micas than to the octaphyllites. Noteworthy is the fact that the deficiencies in the Al,Fe,Li (octahedral) position have been reported. Although Lemke, et al, (1952) report the occurrence at Amelia, Virginia, of many books intermediate between zinnwaldite and muscovite with some gradations from muscovite centers to zinnwaldite borders, our examination of this material shows only a very narrow gradational zone, if any. 201

TABLE L NEW STRUCTURAL STUDIES OF ZINNWALDITE Identification PolymorphicLocalityRemar 2V* Color Locality Remarks Number Form 778 1-layer Medium Green Bohemia Borrowed from Harvard 779 1-layer ND Pale green- Zinnwald Yale No. 1504 brown 1329 1-layer Large Amber brown Bohemia Ontario Museum 1330 3-layer ND Dark brown Altenberg Rabenglimmer, Ontario Museum 1331 1-layer ND Green-black Bohemia Ontario Museum 780 1-layer ND Green-black Fredrickstown, American Museum Mo. 1128 1-layer ND Rust brown Martha Mine, U. S. Nat'l, Pon'gb Bolivia Museum R8089 1251 2-layer ND Pale brown Ubini, W. Murry and Australia Chapman (1931) 1324 3-layer ND Dark brown Brown Derby, Colorado 1337 (a) 3-layer Small Brown Brown Derby, Core zone Colorado 337 (b) 3-layer Small Dark brown Brown Derby, Margin Zone Colorado 786 1-layer Medium Pale brown Amelia, Va. Harvard 787 (a) 3-layer ND Pale brown Amelia, Va. Analyzed by Glass (1935) 1242 (a) 1-layer Medium Brown Amelia, Va. outer edge zone 1242 (b) 3-layer Small Dark brown Amelia, Va. inner zone 1389 (a) 3-layer Small Brown Amelia, Va. Margin Zone; Lemke et al., (1952) *Large 2V means greater than 40~ Medium 2V means 15~ -40~ Small 2V means 0~ -15~ ND means not determined but assumed to be Medium 202

........ sss.....;;.......... Fig. 15. 0-level a-axis Weissenberg000000000000000t photograph000000000 of' 00 0 0 00 0 00 0 0 0 0 0 0 t 0 0 00.............. zinnwaldite wi0; 0 th 2.' B' -layer monoclinic structure':B.' 0 00 00000 Compare: with:Figs. 5. a 6 d t 1X: -000:; t: J;~~~~~~~~~~~~~~......;t~s! B0'0000000000000000000000:0000 00:0 0 te;:000:;:t:; 0 t:;0;0;00:: X 0; 0 0:: X 0 0:0::::ffffffffft00000000j000000:0000000000..........0 i ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~.......B...EESEEEEEE~::...............::: i:*::::::: i:Ei i:EigsiE 00000i000i000 0 ~i0 ~~~~iiiii 0 j t- S.00 Bs - N "BSBS —S,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~......: 0. 000000000000000000 0 0 0 0 0 0 00 00 0 t; 0 0 0;00 i: 0 0 0 0 t;; 0; 0 ti 0000 0 *; 0 00 0 00 0 0 t00 0 0 0 0 0 0 0;00 0 00 000 0 00000000000t............0 B; 0 0:0 * /';. <-: S.,.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...... Fig. 15. O-level a-axis Weissenberg photograph of~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~......... Compare~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.......i s..nd 6

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Murray and Chapman (1931) reported only a partial analysis (the alkalis) of the Ubini material. The Li20 content of 1.15 percent is extremely low for a zinnwaldite and further studies may show that we are not dealing with a true zinnwaldite. Neverthelesso the color is very similar to other zinmwaldites and from the -darkening of the Weissenberg films it is assumed that the Fe content is considerable; certainly more than high Fe, normal muscovites. The itensities of the (o6) reflections with Q odd (Fig. 13; compare with Figs. 5 and 6) from the specimen of zinnwaldite from TJbini (No. 1251) are of an order of magnitude somewhere intermediate between normal muscovite and lithian muscovite (apparently closer to normal muscovite). The intensities of these reflections and other selected (Oke) reflections are listed in Table LI. TABLE LI APPROXIMATE OBSERVED INTENSITIES OF SOME (Oki) REFLECTIONS OF 2-IAYER FOEMS Plane Normal Muscovite 2-L-Lar Zinn- Lithian _ L__ Mu s cI_____wal ite Muscovite 020 W W- a 022 MW MW VW 026- a VW VW 045 a VTVW VW 061 w W- V 065 vw aw vvw 065 VW VW VVW 066 w w vw o67 VW'VW W 069 w Vw C Opt ical Propertie s In Table LII are listed the optical constants of zinnwaldites as reported in the literature. The range of values is as follows: x= 1 e.535 - 1.558 Axial plane: parallel with (010) p = 1.570 - 1.589 Color: gray to amber-brown Y = 1.572 - 1.590 Sp. Gravity: 2.916 - 3.018 2V = 0~ - 70 -- Average 25~ - 30~ 204

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Unfortunately only three analyzed zinnwaldites have precise optical determinations; this prevents any definite correlation of optical data with composition. There does, however, appear to be a general increase in the indices with an increase in iron content. D. Zoning and Overgrowths Baumhauer (1879) reported a zoned zinnwaldite from Zinnwald, BIohemia. According to his descriptions, crystals of zinnwaldite showed zones of different widths normal to the A structure. These zones were yellow, whereas the rest of the crystal was colorless to gray. He postulated that the zoning was due to chemical differences. Upon examination under polarized light, the various zones were found by him to have the same absorption formulas. Sollas (1889) reports a zinnwaldite found in the granite of Mourne Mountains, Ireland, which shows a zonal structure with a dark-green center having an axial angle of 44~04' and a border zone having an angle of 52006'. Sterrett (1923, p. 133) reports zoning in a zinnwaldite from the Palermo Mine, Grafton County, New Hampshire: "The sheets of this mica show a clear brown core and a greenish-blue exterior about the color of indicolite tourmaline. These colors are arranged parallel to planes of crystallization, and the blue contains thin zonal growths of the brown." A zoned crystal from Amelia, Virginia (spec. 1242) upon examination showed optical and structural variations. The dark-brown inner zone had a 2V of 3 to 4~0, p = 1.584, and Y = 1.586, and had crystallized as the 3-layer form. The marginal zone had a 2V of 28~, P = 1.583; and r = 1.588, and crystallized as the 1-layer form. It has been shown in the lepidolites that the 3-layer and 1-layer forms have almost identical chemical compositions; it is believed that the 3-layer form results from a twinning of the 1-layer form. This twinning also probably accounts for the structural variation in this zinnwaldite, as the chemistry of the two zones probably varies very little. A systematic X-ray study of several zoned zinnwaldites from the Brown Derby, Colorado, pegmatite has shown that polymorphic variation does not necessarily exist within markedly zoned crystals. These micas have a light outer zone and a dark core. Both zones have the same optical constants; = 1.584, t = 1.585, and 2V = 0 - 2~, and both have crystallized as the 1-layer type. Lemke et al. (1952 p. 129) report that many books intermediate between zinnwaldite and muscovite occur in the Amelia District of Virginia and that some grade from muscovite centers to zinnwaldite borders. They conclude that the zinnwaldite formed later than muscovite and in some books formed rims around cores of muscovite. 205

TABLE LII OPTICAL PROPERTIES OF ZINNWALDITES (From the literature) — ^-i. in.! (i~,,.'' I —--- - ---- ~ —, — ------- -- ---- -' —----- i'''' -- - Locality a p Axial Color References Zinnad Bohemia 1.1 1.78 1.81 2V = 29-0~ Brole Zinnwald, Bohemia 1.551 1.578 1.581 2V = 29-30~ Brown Larsen and Berman, 1934 Zinnwald, Bohemia 1.5435 1.5729 1.5750 2V = 30-32~ Winchell,1942 -1.5450 -1.5737 -1.5757 Zinnwald, Bohemia 1.541 1.571 1.573 2V = 28047' Light Jakob,1927 Gray Zinnwald, Bohemia 1.5511 1.5777 1.5812 2V = 14~48' Kunitz,1924 2E = 47 30' Zinnwald, Bohemia 1.539 1.564 2V = 30~ Hendricks and Jefferson, 1939 Altenberg, Saxony 1.5572 1.5850 1.5876 2V = 10021' 2E = 36020' Kunitz,1924 Fichtelgebirge, 2E = 47010' Dlirrfeld, 1909 Bavaria Brambach, Saxony 1,572 2V = 25-30~ Hendricks and Jefferson,1939 Volhynia, Russia 1.587 2V = 13-25~ Buryanova,1940 Erongo Schlucht, 1.573 1.576 2V = 31~ + 3~ Frommurze, Gevers, S. W. Africa +0.002 +0,002 and Rossouw,1942 Umberatana, 1.57 2V = 10~ Mawson and South Australia Dallwitz,1945 Wakefield, Quebec 1.5357 1.5596 1.5628 2V = 30-39~ Winchell,1942 -1-.5440 -1.5671 -1.5701 New Ross, Nova 2V = 700 (?) Amber Walker and Parsons, Scotia Brown 1924 Amelia, Virginia 1.550 1.580 1.580 2V = 0-33~ Bronze Glass, 1935 -1.558 -1.589 -1.590 to Gray Amelia, Virginia 1.550 1.584 2V = 0~ Hendricks and Jefferson, 1939 206

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN We have examined some material from Amelia supplied by W. R. Griffitts in order to determine the extent of gradation between zinnwaldite and muscovite. The heavily iron-stained specimen (No. 1389) consists of an incomplete crystal which shows definite zoning. The boundary between the muscovite core and zinnwaldite rim appears megascopically sharp and distinct owing to the brownish tint of the zinnwaldite. Under the microscope the boundary is, in the main, sharp but jagged, as determined by differences in birefringence under crossed nicols. In a few small areas the boundary, on the basis of birefringence and size of the optic angle, appears gradational. However, the gradational zone is narrow, and its lateral extent is limited. The structure of the muscovite is 2-layer(normal). The zinnwaldite has the 3-layer hexagonal structure; the optic angle is small (at the most, 10~) and variable. We have been unable to isolate any of the small "gradational areas" for X-ray studies. In general, however, the specimen represents an overgrowth rather than a gradation. Lemke et al. (1952) have described overgrowths of zinnwaldite on muscovite at the Morefield mine, Amelia, Virginia. Zinnwaldite overgrowths on green biotite are reported by Shibata (1952B) in pegmatites of the Naegi district, Japan, and zinnwaldite overgrowths on muscovite were noted by the same investigator (Shibata, 1952B) occurring in pegmatites at Naegi and Tanokamiyama. 207

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN II. OCCURRENCE Zinnwaldite is found chiefly in granite pegmatites and cassiterite veins associated with greisen. Table LIII lists the localities and type of deposit for most of the zinnwaldite occurrences described in the literature. The survey of the literature describing the occurrences of zinnwaldite leads to the following generalizations concerning the paragenesis of z innwaldite: 1. Zinnwaldite is associated with granitic magmas, especially with pegmatites, greisens, and granites. 2. Rocks containing zinnwaldite usually contain one or more of the following minerals: topaz, cassiterite, lepidolite, feldspar (cleavelandite), beryl, tourmaline, tantalite, columbite, monazite, spodumene, and fluorite. 3. Zinnwaldites in pegmatites are usually associated with Na-Li replacement units. 4. Zinnwaldite probably crystallizes after muscovite and/or biotite but before lepidolite (Shibata, 1952B). 208

TABLE LIII OCCURRENCES OF Z INNWALDITE Location Type deposit Associated minerals Reference Remarks Zinnwald, greisen cassiterite, topaz, Beck (1905) Similarly at Altengerg Bohemia albite Konta (1950) (as rabenglimmer) and Erzgebrige Bavaria pegmatite quartz, albite, Mieleitner At Fuchsbau Forest Germany orthoclase (1921) Cornwall, greisen cassiterite, tbpaz, Dana (1892) Also at St. Just ) England albite Makonu, Rhodesia greisen Zealley (1917) Hunan Prov., China greisen cassiterite, lepidolite Ming and Chang (1935) Italy Baveno granite siderophyllite, lepidolite Gallitelli (1936) York Region, greisen cassiterite, fluorite, Collier (1905) Alaska topaz Umbertarna, pegmatite Mawson and Dallwitz So Australia (1945) Tourmaline Hill, greisen feldspar, quartz Mawson and Dallwitz S. Australia (1945) Ubini, Western pegmatite lepidolite Murray and Chapman Also other localities Australia (1931) in Western Australia Amelia, pegmatite topaz, albite, beryl Glass (1935) Virginia Lemke et al. (1952)

TABLE LIII (CONT.) Locations t Associated minerals Reference Remarks deposit Reference Remarks Brown Derby, pegmatite albite, quartz, Heinrich and Colorado tourmaline Levinson (19535) White Picacho pegmatite lepidolite, Jahns (1952) dist., Arizona spodumene Rockport, Mass. granite quartz, feldspar Cooke (1867) Also Cape Ann Pointe du Bois, pegmatite lepidolite, Heinrich and Manitoba spodumene Levinson (1953) -Karibib, S. W. pegmatite feldspar, tantalite, Frommurze Similarly at other H Africa tourmaline et al. (1942) localities: Omaruru Stavoren Tin greisen cassiterite Wagner (1921) Fields, S. Africa Greenland pegmatite albite, aegirite Flink (1902) At Narsarsuk and Dana (1892) Kangerdluarsuk Russia pegmatite topaz Buryanova (1940) At Volhynia, Lavitinsky, and Turkestan Mt. Range Tordal, pegmatite amazonite, topaz, Oftedal (1942) Norway fluorite, cassiterite Madagascar pegmatite lepidolite Lacroix (1922) At Antsongombato, Antaboaka, etc. Japan-Korea pegmatite lepidolite, spodumene, Shibata (1952B) cassiterite, fluorite, Harada (1936) topaz limori and Yoshimura (1926)

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN PART VII. MISCELLAIEOUS MICAS I. PARAGONITE (NaK)2A14(Si6Al2)020(OH)4; chemically very similar to muscovite except with Na in place of K; K in very subordinate amounts, Schaller and Stevens (1941) have pointed out that the series muscovite-paragonite is not completely represented in nature. They believe there is no longer any doubt as to the validity of paragonite as a separate species even though designation of many alleged soda micas as paragonite was apparently based on faulty analytical determinations. As the optical properties of paragonite, Winchell and Winchell (1951) give: a = 1.564 - 1.577 P = 1.599 - 1.605 = 1.600 - 1.609 2V = ca. 40~ Paragonite cannot be distinguished from muscovite on the basis of optical properties alone. As for its structure, Bannister (1943) has determined the following cell dimensions for paragonite which indicate the close similarity to the normal 2-layer muscovite structure: a = 5.12A b = 8.87A c sinp = 18.95A On synthetic material, Gruner (1942) found: bo = 8.90A co = 19.33 Gruner (1942) suggests that paragonite is uncommon because Na ions are so small that they will not fit readily into the K positions of the micas. In twelvefold coordination positions, bonding with Na is extremely weak. Paragonite is reported as occuring in schists and phyllites, in some with kyanite, also as an alteration of corundum. Jahns (personal communication) has found paragonite in a pegmatite. 211

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN II. ROSCOELITE Roscoelite has been generally accepted as a vanadiferous muscovite in which V has replaced much of the Al. This conclusion was reached by Clarke on the basis of the chemical analyses of Hillebrand (Hillebrand, Turner, and Clarke, 1899). A detailed study of the optical properties of roscoelite was made by Wright (1914), it which the similarities to muscovite, such as the position of the optic plane normal: to (010), were noted. Indices, however, are high: a = 1.610 f = 1.685 and ~ = 1.704 (+ 0.003, sodium light). 2E = 42~ to 69~. Pleochroism is fair3vstrong- y = green-brown, $ = olive-green and a = olive-green. Absorption, y> >(X. Atomic ratios calculated by Wells and Brannock (1946) on the basis of ten chemical analyses show that V and Al combined (with the exception of the Al required to fill tetrahedral positions) total two atoms with octahedral coordination in the mica sheet. This is the theoretical number for a heptaphyllite (muscovite-type) mica. The isomorphous replacement also seems probable on the basis of ionic radii: A13+ = 0.571X;, V3+ = 0.65skx The first published results of X-ray work on roscoelite that we have been able to find are embodied in a statement in Fischer (1942, pp. 376377): "The principal vanadium mineral is micaceous, and it has heretofore been thought to be roscoelite, the vanadium mica; but as a result of recent X-ray studies by Sterling Hendricks of the Department of Agriculture, it is now thought to belong to the imperfectly understood hydrous-mica group of clay minerals," This tentative statement apparently has been widely accepted and the micaceous vanadium mineral of the Colorado Plateau area has subsequently been referred to as vanadium hydromica upon several occasions by geologists studying these deposits. The extent of its dissemination may be gauged by its appearance in a Russian article (Kultiassov and Dubinkina, 1946). Since the results of our studies on roscoelite were not in agreement with this identification, we wrote to Dr. Hendricks for further details. In his reply he states: "I probably made an error in the roscoelite. The sample used was an extremely small one (actually too small to work with) and the powder diffraction photographs were poor. The statement was based on an apparent value of d(001) in excess of 10.OA." The only other reported structural work is a statement in Wells and Brannock (1946, p. 121) attributed to Dr. J. M. Axelrod, which indicates that 212

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN X-ray powder patterns of roscoelite have "... lines very close to those of muscovite, paragonite, and biotite." However, studies by Fischer (1937, P. 929) and by Alice D. Weeks (personal communication) seem to indicate the presence of another vanadiferous micaceous mineral, possibly of chloritic nature, in the V ores of the Placerville district of the Colorado Plateau province. In the University of Michigan Mineralogical Collections only a few fine-grained samples of roscoelite were available, so that the study was begun by means of the powder method. Powder photographs were taken of three specimens from California localities. All of these have spacings that correspond closely to those of the 1-layered monoclinic mica polymorph (Table LIV, column 3); they are slightly, but consistantly, larger than those of the 1-layer lepidolite polymorph (p.141). Specimens of type roscoelite from the Stockslager mine were obtained from the U. S. National Museum and Harvard Mineralogical Museum. Powder photographs of this type material also gave the 1-layer polymorph pattern. On a single flake suitable for Weissenberg study from this locality, a 0-level a-axis photograph indicated the presence of the I-layered monoclinic structure, with the optic plane parallel with the side pinacoid. Samples of two recently analyzed roscoelites, one from Paradox Valley, Colorado, described by Wells and Brannock (1946), and the other from the Placerville District, Colorado, described by Fischer, Haff, and Rcninger (1947) were obtained from Dr. Axelrod and are the same ones on which he based his above cited statement, in Wells and Brannock (1946, p. 121). Powder photographs of these specimens likewise gave 1-layer patterns with no values of d(001) in excess of 10.OA. Two specimens of alleged roscoelite-bearing ore were loaned by Prof. E. N. Goddard for study. One is from the King Wilhelm mine near Jamestown, Colorado; the other from the Buena mine in the same district. Lovering and Goddard (1950, p. 264) report for the Jamestown district that, "Associated with the ore in the Rip Van Dam, King Wilhelm, and Gladiator veins are small amounts of roscoelite..." Roscoelite in this district also has been noted by Bray (1942, p. 787)1 who correlates the presence of this mica in veins with the relatively large amounts of V in the tertiary granodiorite to which the veins are genetically related. Genth (1876) has described and analyzed roscoelite from the nearby Magnolia district in Boulder County, Colorado. The Jamestown specimens consist of thin vein coatings of quartz, fluorite, pyrite, tellurides, and a very fine-grained gray-green m4ca, most of which is intimately intergrown with fine-grained quartz. Single flakes, about 1 mm in diameter, were separated from the specimen from the King Wilhelm mine, and micaceous powders were obtained from both specimens. Weissenberg photographs establish that the single flakes have the 2-layer monoclinic structure of normal muscovite. The powder X-ray photographs show strong quartz contamination, and the polymorphic type is not determinable. Mrs. Alice S. Corey of our laboratory could not substantiate the presence of V in these micas by means of the test described by Axelrod (1946). 213

TABLE LIV X-RAY POWDER DATA FOR ROSCOELITE AND BARIUM MUSCOVITE Muscovite Roscoelite Barium Muscovite Nagelschmidt (1937) California Franklin, N. J. Cu Ka,; X = 1.53736 kCu Kc, X = 1.53736kX 9.98 s 9.91 s 9.80 m 1,635 vw 5.00 s 4.57 m 6.56 m 1.586 vvw 4.49 s 4.38 w 5.70 vw 1.569 vw 3.91 w 4.15 w 4.91 vw 1.530 vvw 3.73 w 3.66 m 4.56 w 1.516 vvw 3.50 m 3.34 s 4.43 w 1.497 w.33 vs 3.12 m 4.20 vvw 1.454 vvw 3.20 m 2.90 vw 3.76 m. 1.419 vvw 3.00 A 2.70 w 3.60 w 1.383 vw 2.88 m 2.60 m 3.52 w 1.356 vvw 2.80 m 2.426 w 3.42 m 1.340 vw 2.57 vs 2.355 vw 3.31 s 1.208 vvw 2.286 vvw 3.23 m 1.193 vw 2.475 w br 2.221 vvw 2.99 m 1.153 vvw 2.167 vw 2.89 m 1.137 vvw 2.5^5 m2.041 vvw 2.8v vvw 1.107 vvw 2.28 w 2.007 vw 2.76 m 1.078 vvw 2.19J 1.665 vw 2.66 vvw 1.056 vvw 2.134 s 1.603 vvw 2.56 s 1.995 vs 1.524 m 2.487 vvw 2.422 vvw 1.730 vw 2.376 vvw 1.651 s br 2.314 vw 1.546 vw 2.226 vvw 1.523 vw 2.165 w 1.500 s 2.123 vw 1.450 vw 2.056 vvw 1.421 vw 1.990 m 1.356 m 1.937 vvw 1.337J 1.848 vvw 1.296 m 1.795 m 1.247 m 1.669 vvw 214

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Roscoelite has been found in the ores of Cripple Creek (Lindgren and Ransome, 1906, p. 128) also associated with quartz, fluorite, and calaveriteo Unfortunately, no analyses of this material are available. Several other occurrences of roscoelite or vanadium-bearing muscovite have been reported. The vanadiferous muscovite (1.08 percent V203) from Schmiedefeld, Saalfeld in Thuringia, Germany, (Jung, 1937) has the normal 2-layer monoclinic structure, as is shown by his X-ray powder spacings (p. 40). It occurs with pyrite in quartz veins in an iron deposito Chatterjee (1931) reports roscoelite in quartz veins with tourmaline and rutile and in schists containing muscovite, rutile, roscoelite, and kyanite in the Bhandara district of India. He states (p. 297), "The roscoelite mica occurs in the kyanitebearing rocks in small nests. It is pale olive-green and slightly pleochroic; and its optical properties correspond to those of muscovite. Tests for nickel and chromium failed, while vanadium was found to be present This mineral is also found in certain quartz veins." Efforts to obtain specimens of roscoelite from this locality were unsuccessful, Service (1943) has described a green vanadiferous mica (1.7 percent V205) from cavities of the Nsuta manganese deposits of the Gold Coast and another green mica (0.62 percent V203) from a quartz vein in the same deposit. Again an attempt to secure some of this material was unsuccessful. The vanadium mica of the vanadium deposits of the Kara-tau Mountains in Kazak, U.S.S.R. has been described as roscoelite by Koslov and Sokolov (1944) and by Turin (1944) and later described as vanadium-oellacherite by Kultiassov and Dubinkina (1946). But as we have shown under oellacherite, the homogeneity and identity of this material are doubtful, Other reported localities for roscoelite include Kalgoorlie, Western Australia (Lindgren, 19335 p. 680) also with native gold and tellurides,. and west of Pretoria, Transvaal, in magnetite ores (Wagner, 1928, po 95-96) of the Daspoort horizon, where it lies within the contact aureole of the Bushveld Complex. Apparently the normal 2-layer monoclinic muscovite structure can accommodate the substitution of significant amounts of V for Al without any detectable change, Those micas with the 1-layer roscoelite structure contain 12 percent V205 or more. We believe that this structural change accompanies the large-scale substitution of V for Al. The few available analyses seem to indicate that there is no complete series between vanadian muscovite and roscoelite. It would perhaps be instructive to determine the structure of the vanadiferous mica from the Magnolia district, Colorado (Genth, 1876), which contains 7.78 percent V203o Since roscoelite is both chemically and structurally distinct from other micas, it can no longer be regarded as a vanadiferous muscovite, but must be given the status of an independent species. ____________________________________ 215

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN III. TAENIOLITE Taeniolite is an extremely rare member of the mica group whose formula is K2Mg4Li2Si8020F4- It is characterized by the absence of essential aluminum, and for this reason is unique among the micas. Mg and Li alone appear to occupy octahedral positions. The high SiO2 and F along with a substantial Li2O (approximately 3.5 percent) content make it appear related to octophyllite lepidolite; future investigations may show this mineral to be a magnesian lepidolite in which Mg has replaced Al. An occurrence of taeniolite from Magnet Cove, Arkansas, is described by Miser and Stevens (1938). They also discuss the relation of taeniolite to other micas. The optical properties of the Arkansas material are: a = 1.522 1 = 5 = 1.553 -o- = 0.031j 2V = near 0~ Hendricks and Jefferson (1939) demonstrated that taeniolite from this locality has the 1-layer monoclinic structure. Taeniolite is found in veins and pockets with quartz, dickite, rutile and other minerals at this locality and also at Narsarsuk, Greenland, and also possibly in the Kola Peninsula. 216

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN PART VIII. PARAGENESIS OF PEGMATITIC MICAS I. SEQUENCE OF CRYSTALLIZATION OF MICAS A. Zonal Growths In the muscovites, the common type of color zoning consists of darker-brown cores and lighter-brown or greenish-brown outer zones. From the relation of color to composition, this indicates a decrease in Fe3 and Ti outward and probably an increase in Fe2 outward. Analyses 8a and 8b, Table LV, representing a zoned muscovite from the Jasper mine are typical. 8a 8b green margin ruby core Total Fe as Fe203 2.9 2.4 MgO.45.75 TiO2.05.30 The inner parts of muscovite crystals that show zoning normally are richer in hematite indicating the greater availability of Fe3 and the greater ability of muscovite to incorporate it during the early growth stages. Zoned biotites are less common, but some examples have been analyzed by Shibata (1952A) from pegmatites in the Naegi district, Japan. Green outer zones have somewhat more Fe2, slightly less Fe3, and markedly less Ti than brown cores. green margin dark-brown core Fe2 7.44 7.73 FeO 24.12 22.41 MgO 0.29 0.62 TiO^ 0.08 0.76 217

TABLE LV SPECTROCHEMICAL ANALYSES* OF MUSCOVITES FROM JASPER PEGMATITE (FRANKLIN-SYLVA DISTRICT, N. C.)..6..... 8a. - b ruby green margin ruby core inner part of i art zone zoned,-book outer part of wall zone wall zone Fe'20 3.2 2.9 2.4 MgO 1.7.45.75 MnO.04.07.07 TiO2.32 0O5.30 CaO.014.002.001 SrO.001- - BaO 8.4.07 Na20.67.75.90 Li2O -- RbO -- - Cs2O — Ga203 07 6.017.0162 Sc.0 Sc20.003 5 Co 03.0002 ---- - 0004 Cr203.0003.0001.0003 Sn02.004.004.004 V205.014.001.01 *Spectrochemical analyses by C. Eo Harvey 218

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN B. Overgrowths Between Micas Overgrowths between mica species are commonly of the types: 1. Muscovite overgrowths on biotite. 2. Lepidolite overgrowths on muscovite. 3. Lepidolite overgrowths on biotite. 4. Zinnwaldite overgrowths on muscovite. 5. Zinnwaldite overgrowths on biotite. Reverse relations, such as biotite on muscovites and muscovite on lepidolite are known but are rare for the pairs. C. Localization in Pegmatites Biotites of magmatic origin occur mainly in wall zones or outer intermediate zones. In wall zones they are earlier, normally, than any accompanying muscovite. Wall-zone and intermediate-zone biotite is earlier, of course, than core-margin muscovite. Hydrothermal biotite occurs as fracture fillings but very rarely in replacement units. Where biotite of the fracturefillings type occurs associated with fracture-filling or replacement muscovite, the latter is the younger. Shibata (1952A) has analyzed several paragenetically distinct biotites from the Iisaka pegmatite, Japan: Older Younger brown biotite in green biotite, fracture graphic granite zone controlled, associated with fergusonite Fe 20 12.89 11.72 FeO 20.06 23.73 MgO 1.21 0.14 TiO2 0.81 0.55 219

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN The younger biotite is lower in Fe3,Mg, and Ti but higher in Fe2. Muscovites of wall zones and outer zones are older than those of core-margin units, which in turn, are older than muscovite formed hydrothermally along fractures and by replacement. Table XIX illustrates the differences in wall-zone (ruby) and core-margin (green) muscovites from the Kiser Pegmatite. The younger muscovite has higher total Fe and lower Ti. The muscovites of the Franklin-Sylva district, North Carolina, are almost entirely of magmatic origin, and most of them are wall-zone micas. The micas of the Petaca district, New'Mexico occur chiefly in replacement.' units and are in the main of hydrothermal origin (Jahns, 1946). Table LVI contrasts their average compositions. Again, the chief differences are in Fe,Mg,Mn, and also in Li and F. Na, Ti and some of the trace elements also show variations, Liashchenko (1940) has analyzed two muscovites from the Biriussa mica mine in eastern Siberia (Table LVII). Number 1 is a normal pegmatitic muscovite; No. 2 is a fine-grained secondary muscovite that forms as pseudomorphs replacing the book muscovite (No. 1). The TiO content of No. 2 is abnormal and may be presumed to be due to inclusions of a Ti mineral. It is to be noted that the secondary muscovite is phengitic, is slightly deficient in K, and is very low in Fe3 + Fe2. As early as 1887, Scharizer (1887B) analyzed the several micas from the Schlttenhofen pegmatite.(Table LVIII). He concluded that the youngest is richest in Si02, poorest in divalent bases, and is also the lightest in color. In the outer zone, muscovite of type I also forms overgrowths on biotite so that the complete sequence is: Biotite > muscovite I - muscovite II > lepidoliteo He believed that the type II muscovite is poor in Fe3 because tourmaline formed ith it, which utilized that element. Shibata (1952A) has deciphered the paragenetic sequences of micas in several districts: 220

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 1. Naegi district, Japan younger brown biotite green biotite protolithionite- zinnwaldite - lepidolite muscovite - sericite yellow mica 2. Bunsen pegmatite, South Kanky-do-, Korea younger biotite radial muscovite - zinnwaldite w-hite pink lepidolite lepidolite 3. Sakihama, Japan See Table LIX. Heinrich and Levinson (1953) have cited evidence to show that in many cases rose muscovite is younger than lepidolite. 221

TABLE LVI COMPARISON OF AVERAGE COMPOSITION* OF MUSCOVITES FROM FRANKLIN-SYLVA, N. C. AND PETACA, N. MEX,, DISTRICTS F aEiklEh-Sylv-a Petaca North Carolina New Mexico Fe203 2.77 4.24 MgO 1.12.12 MnO.06.47 TiO2.17.105 CaO.015.005 SrO.0009 absent BaO.19.oo6 Na20.83 1.08 Li2O absent.46 Ga203.017.041 Sc20O.0014.0002 Co203.0002.0001 Cr203.0002.00001 SnO2.0018.014 V205.0066.0004 F.029 1.00 ABased on spectrochemical data by C. E. Harvey 222

TABLE LVII TWO MUSCOVITES ANALYZED BY LIASHCHENKO (1940) 1i 2 Si02 44o01 51.07 Ti02 0.57 4.52 ZrO2 -- 0.03 Al2 0 34.25 29.29 Cr20 n.d. --- Fe20O 1.09 0.45 FeO -- 0.20 MgO 1.35 1.36 CaO 0.79 K20 10.45 7.71 Na 0 0.38 --- H26+ 5.78 5.24 H O- 1.62 0.56 Cb2 0.18 0.16 100.45 100.59 TABLE LVIII MICAS FROM THE SCHUTTENHOFEN PEGMATITE ANALYZED BY SCHARIZER (1887B) I II III muscovite, muscovite lepidolite brown, oldest yellow youngest SiO 43.67 44.08 49.25 A1203 36.69 36.83 25.26 Fe203 2.10 0.48 FeO 0.55 0.74 0.84 MnO tr 0.25 0.85 K20 8.57* 11.10 13.85* Na20 1.95 0.20 0.35 Li 0 tr 0.37 5.38 Ca6 -- 0.20 H20+ 4.35 4.98 1.76 H20- 1.15 1.17 -- F 0.35 0.19 5.68 99.38 100.59 103.22 0 = F = 0,14 o.o8 2.38 99.24 100.51 100.84 *Includes Rb20O and Cs20 223

TABLE LIX TWO MICAS FROM SAKIHAMA, JAPAN ANALYZED BY SHIBATA (1952A) Zinnwaldite Lepidolite (core margin) (Core, in vugs) Si02 45.98 46.86 TiO2 0.22 0.14 A1203 18.04 21.25 Fe20 3.56 1.01 FeO 6.00 0.16 MgO 1.50 0.52 MnO 2.46 7.55 CaO 1.80 0.78 Na2O 0.25 0.22 K20 9.91 9.48 Li20 4.57 5.20 Rb2o --- 1.36 Cs2O -- 0.04 H20+ 0.86 0.25 2 H20.- 1.14 1.28 F 5.66 8.31 102.08 1t4.;41 0 F 2.38 3.66 99.70 100.75 224

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN II. CHEMICAL EVOLUTION Shibata (1952B) concludes that there exist three lines of descent in the evolution of pegmatitic micas: 1. biotite - lepidomelane (Mg-Fe2 > Fe2-Fe3) 2. muscovite ) lithia mica 3. biotite -- lithia mica He also notes (p. 160) that, "In later stage minerals found in pegmatites the ratio of MnO to FeO increases and at the same time lithium increases especially in mica, etc. When magnesium almost diminishes in later stage-minerals, manganese takes its place. Relative stage of magmatic differentiation is represented by the ratio- of FeO/(MgO + FeO) at the earlier stage and by MnO/(FeO + MnO) at the later one." With these general conclusion the writers are in agreement, as was expressed by Heinrich (1952). Shibata also states (p. 160), "Ferric iron can be derived from ferrous Oone by oxidation by pegmatitic gases and so the ratio of ferric to ferrous iron has no constant relation, otherwise ferric iron increases in minerals of the later stage at the expense of ferrous iron." The writers agree that ferrous iron can be converted to ferric by means of the escape of pegmatitic gases, but conclude that such oxidation is more likely to take place in the outer parts (wall zones) of pegmatites. Thus wall-zone muscovites contain more Fe3 (and exsolved Fe20 ) than core-margin types and biotites of outer zones will also probably be shown to contain more Fe3 than core-margin types (most biotites included in green core-margin muscovites are themselves green) and more Fe3 than fracture-filling biotite. Table LX summarizes the average contents of the various oxides in the micas analyzed for this study. In the muscovites the chemical evolution normally involves the changes indicated in Table LXI. Normally the sequence muscovite - lithian muscovite -,lepidolite is accompanied by: -____________________________________ 225 _

TABLE LX AVERAGE CONTENTS* OF VARIOUS OXIDES IN MICAS Muscovite Lep idolite Rose Muscovite Fe203 2.90 0.48 0.11 MgO 0.93 0.04 0o04 MnO 0.11 1.16 0.33 TiO2 0.13 0.01 0.002 CaO 0.007 0.007 0.004 SrO 0.0005 o.oo6 0.0004 BaO 0.114 0.0003.oo0006 Na2O 1.02 0.41 0.87 Li2O 0.09 3.77 0.32 Rb2O - 0.54 0.62 CsO o.o8 0.11 Ga203 0.022 0.017 0.024 Sc2 0.0011 Oo0006 Co20g 0.0002 -- 0.00oo04 Cr20 0.0001 0.000004 0.00004 SnO2 0.0067 0.021 0.025 V2 0.000oo4 0.0001 0.00005 F 0.24 5.41 0.42 *Based primarily on spectrochemical analyses by C., E. Harvey 226

TABLE LXI CHEMICAL EVOLUTION OF MUSCOVITE Oldest Youngest Outer- one Core-margin Replacement Fine-grained muscovite muscovite muscovite replacement muscovite.___..____________,___......__( "sericite") Fe3 maximum decreasing Fe2 - increasing -- maximum variable gener-,i'' ____ _ ally low Mg maximum decreasing I Mn increasing - - maximum' high Ti maximum - generally decreasing 9 Ca - decreasing slightly _ 0 Ba maximum - decreasing, Na increasing --- > maximum slight increase F t, lolow maximum low and variable F i | g increasing high \ slight 0 lowest; generally OH I decrease; high Si |generally the same marked increase maximum K slight increase great increase lowest Ca, Sr, Co, Cr, Sn show little systematic variation; V, Sc appear to be concentrated slightly in the older micas; Ga is slightly higher in younger micas; Rb and Cs are absent in most micas except the rose types. 227

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 1. An increase in LiMnSr,Rb,Cs,F. 2. A decrease in Mg, total Fe,Ti,Na,Ba. 3. A slight decrease in V(?)C,CrCo,Sc. 4. A slight increase in Sn(?). 5. Essentially no change in Ca,Ga. Rose muscovites, which are generally later than lepidolites, differ from the latter in having still lower total FeMnTi,SrLiSc, and F; and higher Na and perhaps Ga. Rb,Cs,Ca,BaSn, and V are essentially the same. These appear to be the only muscovites that normally contain Rb and Cs. In the sequence:zonal biotite- fracture-filling biotite the data are relatively few, but the indicated accompanying changes are: 1. Decrease in Fe3,Ti,OH(?). 2. Increase in Fe2,Li(?),F(?) The chemical changes that appear in successively younger generations of pegmatitic micas reflect in general, and in detail, the differentiation trends of pegmatitic mgmas and their derivative hydrothermal solutions as well as the course of differentiation of magmas in general. In the magmatic micas with decreasing temperature of crystallization Mg is supplanted by Fe2, and the Fe2 by Mn. The relatively high Fe3 contents of older (higher-temperature) micas may be due to oxidation by gases escaping from the partly crystallized pegmatite magma. With decreasing temperatures less Ti and Ba can be incorporated in micas. LiRbCs, most Mn, and F donot enter appreciably into the structure of the magmatic micas but are rejected to the hydrothermal fraction, along with additional Si. 228

REFERENCES Amelinckx, S. (1952A), Nature, 169, p. 580. (1952B), Natuurwet. Tijdschr. (Belgium), 34, pp. 73-75. Axelrod, J. M. (1946), U.S.G.S. Bull., 950, pp. 19-23. and Grimaldi, F. S. (1949), Am. Mineral., 34, pp. 559572. Bannister, F. A. (1943), Min. Mag., 26, pp. 304-307. Bastin, E. S. (1911), U.S.G.S. Bull., 445, p. 50. Bauer, L. H., and Berman, H. (1933), Am. Mineral., 18, p. 30. Baumhauer, H. (1879), Zieit. Krist., 3, pp. 113-121. (1900), Zeit. Krist., 32, pp. 164-176. - (1903), Eclogae Geol. Helvetiae, 7, p. 354. (1912), Zeit. Krist., 51, pp. 344-357. Beck, R. (1905), The Nature of Ore Deposits: p. 205. Berggren, T. (1940), Geol. Foren. Forh., 62, pp. 182-193. (1941), Geol. Fo5ren. Forh., 63, pp. 262-278. Berman, H. (1937), Am. Mineral., 22, pp. 342-408. Bowman, H. L. (1902), Min, Mag., 13, p. 98. Bragg, W. L. (1937), Atomic Structure of Minerals: Cornell Univ. Press, Ithaca, N. Y. Brammall, A.,Leach, J. G. Co, and Bannister, F. A. (1937), Min. Mag., 24, pp. 507-520. Bray, J. M. (1942), Geolo Soc. Amer. Bull., 53, pp. 765-814. Brigger, W. C. (1921), Die Eruptivgesteine des Cristianiagebietes: IV, p. 294. 229

Brooke, H. J., and Miller, W. H. (1852), An Elementary Introduction to Mineralogy by the Late William Phillips: London. Brown, G., and Norrish, K. (1952), Min. Mag., 29, pp. 929-932. Buryanova, E. Z. (1940), Mem. Soc. Russe Min., 69 (4), pp. 519-540. Cameron,; E. N., Jahns, R. H., McNair, A. H., and Page, L. R. (1949), Econ. Geol., Monograph 2. Chapman, R. W., and Williams, C. R. (1935), Am. Mineral., 20, pp. 502530. Chatterjee, S. K. (1931), India Geol. Surv. Records, 65, pp. 285-305. Chester, A. H. (1894), Trans. N. Y. Academy of Sciences, 15, p. 181. (1896), A Dictionary of the Names of Minerals: New York. Chudaba, K. (1929), Cent. Min., Abt. A, pp. 11-18. Clark, R. W., and Hunt, W. F. (1915), Cent. Min., pp. 666-668. Clarke, F. W. (1887), Amer. Journ. Sci., 34, pp. 131-137. (1888), U.S.G.S. Bull., 42, pp. 11-27. (1889), Amer. Journ. Sci., 138, pp. 384-393. (1914), U.S.G.S. Bull., 588. Coats, R. R., and Fahey, J. J. (1944), Am. Mineral., 29, pp. 373-377. Collier, A. J. (1905), U.S.G.So Bull., 229, p. 20, Collins, R. S. (1942), Min. Mag., 26, pp. 254-259. Cooke, J. P. (1867), Amer. Journ. Sci., 2nd Series, 43, pp. 217-230. Cross, W. (1897), Amer. Journ. Sci., Ser. 4, 4, pp. 115-141. Daimon, N. (1952), Jap. Journ. Ind. Chem., 55, pp. 694-695. Dana, E. S. (1892), System of Mineralogy: 6th ed., New York. (1922), A Textbook of Mineralogy: 3rd ed., New York. (1926), A Textbook of Mineralogy: 3rd ed., New York. Dana, J. D. (1854), System of Mineralogy: 4th ed., New York. (1867), Amer. Journ. Sci., 94, pp. 252-263. 230

Dana, J. D. (1868), System of Mineralogy: 5th ed., New York. (1875), System of Mineralogy: 5th ed., New York. (1882), System of Mineralogy: 5th ed., New York. Daub, R. (1913), Neues Jahrb, (1), pp, 214-217. Des Cloizeaux, A. (1862), Manuel de Mineralogie: Paris. Doelter, C. (1914), Handbuch der Mineralchemie: 2, Part 2, p. 442. Durrfeld, V, (1909), Zeit. Krist,, 46, pp. 563-598. Ellsworth, H. V, (1932), Canada Geol. Survey, Econ. Geol. Ser., No. 11. Erametsa, 0O. Sahama, T. G., and Kanula, V. (1943), Bull. Geol. Comm. Finlande, 128, pp. 80-88. Eskola, P. (1914), Bull. Geol. Commo Finlande, No. 40, p. 37. Fermor, L. L. (1909), Mem. Geol. Surv. India, 37. Fischer, R. P. (1937), Econ. Geol., 32, pp. 909-951. (1942), U.S.G.S. Bull.,, 936 P. Haff, J. C., and Romiger, J. F. (1947), Colorado Scientific Soc. Proc., 15, No, 3, pp. 117-146. Flink, G. (1899), Meddelser om Gronland, 24. (Cit. in Neues Jahrb., Referata, pp. 29-30, 1902). Ford, W. E. (1932), A Textbook of Mineralogy, by E. S. Dana: 4th ed., John Wiley and Sons, Inc., New York. Foster, M. D. (1952), Bull. Geol. Soc, Am., 63, pp. 1250-1251. Frank, F. C, (1949), Discus. of Faraday Soc., No. 5, Crystal Growth, pp. 48-54. Freudenberg, W. (1919), Mitto Badisch. Geol. Landesanst., 8, p. 317. Frommurze, H., Gevers, T. W,, and Rossouw, P. J. (1942), Geol. Survey South Africa, Explanation of Sheet No. 79, pp. 107-114. Frondel, C. (1936), Am. Mineral,, 21, pp. 777-799(1940). Am, Mineral., 25, pp. 69-87. and Ashby, Gq E, (1937), Amn Mineral., 22, pp. 104-121? 231

Furcron, A. S., and Teague, K. H. (1943), Georgia State Div. of Conservation, Geol. Surv. Bull., 48. Gadeke, R. (1938), Chemie der Erde, 11, pp. 592-636. Gallitelli, P. (1936), Peridico Min. Roma, 7, pp. 61-76. Gennaro, V. (1925), Rend, Accad. naz. Lincei, Ser. 6, 2, pp. 508510. Genth, F. A. (1876), Amer. Journ. Sci., 110, pp. 32-35. Ginsburg, J. (1930), Zeit. Praht. Geol., 38, pp. 24-30. Glass, J. (1935), Am. Mineral., 20, pp. 741-768. Grigoriev, D. P. (1935), Mem. Soc. Russe Min., 64, pp. 76-79. ( (1936), Bull. Soc. Nat., Moscow, 17(4-5), pp. 14-30. Grim, R. E., and Bradley, W. F. (1951), The Mineralogical Society, London, pp. 138-154. Bray, R. H., and Bradley, W. F. (1937), Am. Mineral., 22, pp. 813-829. Grout, F. F. (1924), Am. Mineral., 9, pp. 159-165. Gruner, J. W. (1942), Am. Mineral., 27, pp. 131-134. (1948), Am. Mineral., 33, pp. 679-691. Hall, C. E. (1885), Penn. Geol. Survey Report of Progress C5, Part I. Hall, A. L. (1926), Proc. Geol. Soc. of South Africa, 29, pp. 17-46. Hall, A. J. (1941), Am. Mineral., 26, pp. 29-33... —----- (1941), Am. Mineral., 26, pp. 34-41. Hallimond, A. F. (1925), Min. Mag., 20, pp. 305-318. (1926), Min. Mag., 21, pp. 25-33. (1927), Am. Mineral., 12, pp. 413-414. (1947) Min. Mag., 28, pp. 230-242. Hanley, J. B., Heinrich, E. Wm., and Page, L. R. (1950), U.S.G.S., Prof. Paper 227, pp. 29, 40. Harada, Z. (1936), Journ. Fac. Sci. Hokkaido Imp. Univ., Sapporo, Japan, 3, Sero 4, pp. 221-362. 232

Harker, A. (1932), Metamorphism: London. Harry, W. T. (1950), Min. Mag., 29, pp. 142-149. Heinrich, E. Win. (1946), Am. Journ. Sci., 244, pp. 836-848. (1948), Am. Mineral., 33, pp. 420-448; 550-588. (1950), Am. Mineral., 355 pp. 173-184. (1951), Am. Mineral., 36, p. 634. (1952), Bull. Geol. Soc. Amer., 63, p. 1261. (1953), Am. Mineral., 38, pp. 68-87. and Levinson, A. A. (1953), Am. Mineral., 38, pp. 25-49. and Olson) J. C. (1953) U.S.G.S., Prof. Paper 248-G. Hendricks, S. B., and Jefferson, M. E. (1939), Am. Mineral., 24, pp. 729-771~ Hess, F. L., and Fahey, J. J. (1932), Am. Mineral., 17, pp. 173-176.' Hey, M. H. (1950), An Index of Mineral Species and Varieties Arranged Chemically: British Museum, London. Hillebrand, W. F.,Turner, H. W., and Clarke, F. W. (1899), Amer. Journ. Sci., 157, pp. 451-458. Hintze, C. (1897), Handbuch der Mineralogie: 2, pp. 516-656. Leipzig. Holzner, J. (1936), Zeit. Krist., 95, pp. 435-450. Hutton, C. 0. (1940), New Zealand Journ. Sci. Tech., Sec. B, 21, PP. 330-333. (1942), Trans. Roy. Soc. New Zealand, 72, pp. 53-68. and Seeyle, F. T. (1947), Trans. Roy. Soc. New Zealandg 76, ppa 481-4916 Igelstrcm, L. J. (1872), Neues Jahrb., p. 296. Iimori, S., and Yoshimura, S. (1926), Chem. Soc. Japan, 1, pp. 215219. Inoue, T. (1950), Journ. Geol. Soc. Japan, 56, pp. 71-77~ Jackson, W. W., and West, J. (1930), Zeit. Krist., 76, pp. 211-227. 233

Jackson, W. W., and West, J. (1933), Zeit. Krist., 85, pp. 160-164. Jacobson, G., and Webb, J. S. (1946), Geol. Surv. of Nigeria, Bull. 17. Jahns, R. S. (1945), Geol. Soc. Amer. Bull., 56, p. 1170. (1946), New Mexico Bur. Mines Min.Res., Bull. 25. (1952), Bull. Univ. of Arizona, Ariz. Bur. Mines, 23, No. 5 (Min. Tech. Ser. No. 46, Bull. No. 162). and Lancaster, F. W. (1950), U.S.G.S., Prof. Paper 225. and Wright, L. A. (1951), Calif. Div. of Mines, Special Report 7-A. Jakob, J. (1920), Helvetica Chimica Acta, 3, pp. 669-704. - -- (1921), Zeit. Erist., 56, pp. 295-304. (1925A), Zeit. Krist., 61, pp. 155-163. (1925B), Zeit, Krist., 62, pp. 443-453. (1926), Zeit Krist., 64, pp. 430-454. -(1927), Schweiz Min. Petrog. Mitt., 7, pp. 139-141. (1928), Zeit. Krist., 69, pp.217-225. (1929A), Zeit. Krist., 69, pp. 403-410. (1929B), Zeit. Krist., 69, pp. 511-515. (1929C), Zeit. Krist., 70, pp. 493-496. --- -- (1929D), Zeit. Krist., 72, pp. 327-380. ----- (1931), Zeit. Krist., 79, pp. 367-368. (1937), Schweiz Min. Petrog. Mitt., 17, pp. 149-153. (1938), Schweiz Min. Petrog. Mitt., 18, p. 473. -and Parga-Pondal, I. (1932), Zeit, Krist., 82, pp. 271284. Johannsen, A. (1932), A Descriptive Petrography of the Igneous Rocks: II, Univ. of Chicago Press, p. 93. h — - (1937), A Descriptive Petrography of the Igneous Rocks: III, Univ. of Chicago Press, p. 35. 234

Johannsen, A. (1938), A Descriptive Petrography of the Igneous Rocks: IV, Univ. of Chicago Press, p.. Jung, H. (1937), Chem. Erde, 11, pp. 38-44. Kauffman, A. J., Mortimore, D. M., and Hess, H. D. (1950), U. S. Bur. Mines, Dept.of Investig. 4721. Kawano, Y. (1933), Proc. Imp. Acad. Tokyo, 9, pp. 613-616. --- - (1942), Journ. Jap. Assoc. Mineralogists, Petrologists and Econ. Geologists, 27, pp. 283-90. Kennard, T. G., and Howell, D. H. (1941), Am. Mineral., 26, p. 405. Kenngott, A. (1860), Ubersicht der Resulate Mineralogischer Forshungen: Leipzig. Kerr, P. F., and Hamilton, P. K. (1949), Prelim. Report No. 1, Amer. Petrol. Instit. Proj, 49, Clay Mineral Standards, Columbia Univ.) New York. Anderson, T. P., Rasor, C. A., Pill, R. J., Hamilton, P. K., Wright, H., and Cohen, W. J. (1951), United States Atomic Energy Commission, RMO-797. Knopf, A. (1929), U.S.G.S., Prof. Paper 157, p. 38. Kokscharov, N. V. (1875), Materialien zur Mineralogie Russlands: 7, pp. 222-344, St. Petersburg. Konta, J. (1950), Rozpravy Ciski Akad., C1. II, 60, No. 13, 17pp. Koslov, N. A., and Sokolov, V. A. (1944), Akad. Nauk. Kazakstan S.S.R. Izv., 2-3, pp. 82-99. Kultiassov, S. V., and Dubinkina, R. P. (1946), Mem. Soc. Russe Min., 75 (3), pp. 187-192. Kunitz, W. (1924), Neues Jahrb., Beilage-Band, 50, pp. 365-413. Lacroix, A. (1922), Mineralogie of Madagascar: Vol. I, p. 476. Landes, K. K. (1925), Am. Mineral., 10, p. 371. (1938), Am. Mineral., 33, pp. 359-390. Larsen, E. S. (1941), U.S.G.S., Prof. Paper 197-A, p. 53. andBerman, H. (1934), U.S.G.S. Bull., 848, pp. 163237. 235

Lemke, R. W., Jahns, R. H., and Griffitts, W. R. (1952), U.S.G.S., Prof. Paper 248-B. Lester, J. G. (1946), Am. Mineral., 31, p. 77. Levinson, A. A. (1953), Am. Mineral., 38, pp. 88-107. Liaschenko, C. P. (1940), Mem. Soc. Russe Min., 69 (4), pp. 477-479. Linck, G. (1901), Zeit. Krist., 35, pp. 315-320. Lindgren, W. (1933), Mineral Deposits: McGraw-Hill Book Company, Inc., New York. and Ransome, F. L. (1906), U.S.G.S., Prof. Paper 54. Lovering, T. S., and Goddard, E. N. (1950), U.S.G.S., Prof. Paper 223. Lundblad, B. (1942), Geol. Foren. Forh., 64, pp. 55-60. Macgregor, A. M. (1945), Min. Mag., 27, pp. 157-165. Machatschki, F. (1930), Cent. Min., Geol. und Paleon., Abt. A, pp. 255-267. Marginac, C. (1847), Suppl. Bibl. Universe. Geneve, Arch. Sci. Phys. Nat., 6, pp. 293-304. Mauguin, C. (1927), Compt. Rend., 185, pp. 288-291. (1928A), Compt. Rend., 186, pp. 879-881. (1928B), Bull. Soc. Fr. Min., 51, pp. 285-332. Maurice, C. S. (1940), Econ. Geol., 35, pp. 57-67. Mawson, D., and Dallwitz, W. B. (1945), Trans. Roy. Soc. South Australia, 69, pp. 22-49. Meixner, H. (1939), Ann. Naturhist. Museums Wien, 50, pp. 694-703. Merkulova, G. A. (1939), Mem. Soc. Russe Min., 68 (4), pp. 559-561. Mieleitner, K. (1921), Zeit. Krist., 56, pp. 90-94. Ming, H. M., and Chang, K. (1935), Mem. Nat. Research Inst. Geol. Acad. Sinica, Naking, No. 15, pp. 15-72. Miser, H. D., and Stevens, R. E. (1938), Am. Mineral., 23, pp. 104-110. Murdoch, J., and Webb, R. W. (1938), Am. Mineral., 23, pp. 349-355. 236

Murray, D. G., and Chapman, F. E. (1931), Journ. Roy. Soc. West Australia, 17, pp. 151-155. Nagelschmidt, G. (1937), Zeit. Krist., 97, pp. 514-521. Nockolds, S. R., and Mitchell, R. L. (1948), Trans. Roy. Soc. Edinburgh, 61, pp. 533-575. Noll, W. (1932), Naturwissenschaften, 20, p. 283. Odman, O. H. (1938), Geol. Foren. Forh., 60, H. 2. ------ (1950), Sveriges Geol. Undersokning, Ser. C, Arsbok, 44, No. 2. Oftedal, I. (1942), Norsk Geologisk Tidsskrift, 22, pp. 1-14. Olson, J. C. (1942), U.S.G.S. Bull., 931 P, p. 579. Orcel, M. J. (1926), Bull. Soc. Fro Min., 48, pp. 362-366. Pagliani, G. (1937), Atti. Soc. Ital. Sci. Nat. Mus. Civ. Milano, 76, pp. 13-16. ----- (1940), Atti. Soc. Ital. Sci. Nat. Mus. Civ. Milano, 79, pp. 20-22. - (1949), Atti. Soc. Ital. Sci. Nat. Mus. Civ. Milano, 88, pp. 191-198. Palache, C. (1929), Am. Mineral., 14, pp. 43-47. (1935), U.S.G.S., Prof. Paper 180. Pardee, L. T., Larsen, E. S., and Steiger, G. (1921), Journ. Wash. Acid. Sci., 11, No. 2. Park, C. F. (1942), U.S.G.S. Bull., 931-R. Partridge, F. C. (1937), Trans. Geol. Soc. South Africa, 39, pp. 457-460. Pauling, L. (1930), Proc. Nat. Acad. Sci., 16, pp. 123-129. Peacock, M. A., and Ferguson, Ro B. (1943), Univ. of Toronto, Geol. Ser., No. 48, pp. 65-82. Pecora, W. T., Klepper, M. R., and Larrabee, D. M. (1950), U.S.G.S. Bull., 946-C. Pehrman, G. (1945), Acta Acad. Abo. math. physo, 15 (2), p. 57. Penfield, S. L. (1893), Amer. Journ. Sci., 46, pp. 288-295. 237

Pieruccini,. (1950), Atti. Soc. Toscana Sci. Nat. Mem., 57, Ser. A, pp. 145-17 ~ Pogue, J. E. (1911), Zeit. Krist., 49, p. 455. Postel, A. W., and Adelhelm, W. (1944), Am. Mineral., 29, pp. 279-290. Powell, R. W., and Griffiths, E. (1937), Proc. Roy. Soc., Ser. A, 163, pp. 189-198. Prider, R. T. (1940), Min. Mag., 25, pp. 373-387. Prior, G. T. (1908), Min. Mag., 15, p. 385. Rabbitt, J. C. (1948), Am. Mineral., 33, pp. 263-323. Ramaseshan, S. (1945), Proc. Indian Acad. Science, 22, Sec. A, pp. 177-181. Ramberg, H. (1948), Econ. Geol., 43, p. 553. Rammelsberg, C. (1862), Zeit. Deut. Geol. Gesell., 14, pp. 758-764. (1878), K. Akad. Wissen., Berlin, Monatsberichte, pp. 613-631. (1879), K. Akad. Wissen., Berlin, Monatsberichte, pp. 833-853. Ramsdell, L.. (1947), Am. Mineral., 32, pp. 64-82. Rankama, K., and Sahama, T. G. (1950), Geochemistry: Univ. of Chicago Press. Reusch, E. (1869), Pogg. Ann. Phys. Chem., 138, pp. 628-638. Rinne, F. (1925), Zeit. Krist., 61, pp. 113-124. Rowledge, H. P. (1945), Annual Report, Government Mineralogist Analyst and Chemist, Western Australia, for 1943. Roy, R. (1949), Journ. Amer. Ceram. Soc., 32, p. 202. Sandberger, F. (1875), Neues Jahrb. fir Mineralogie, pp. 624-626. Sandraa, A. (1950), Bull. Soc. Fr. Min., 73, pp. 446-453. Savornin, A. (1934), Soc. Fr. Min., Bull. 57, pp. 130-139. Schafhautl, K. F. (1842), Lieb. Ann., 44, p. 40. (1843), Lieb. Ann., 46, p. 325. 238

Schaller, W. T. (1916), U.S.G.S. Bull., 610, pp. 139-140. (1930), Am. Mineral., 15, pp. 566-574. (1950), Min. Mag., 29, pp. 405-415. and Henderson, E. P. (1926), Am. Mineral., 11. pp. 5-16. and Stevens, R. E. (1941), Am. Mineral., 26, pp. 541545. Scharizer, R. (1886), Zeit. Krist., 12, pp. 1-17. (1887A), Zeit. Krist., 13, pp. 15-24. (1887B), Zeit. Krist., 13, pp. 449-473. Scheerer, T. (1862), Zeit. Deutsch. Geol. Gesell., 14, pp. 25-150. Serdiuchenko, D. P. (1948A), Akad. Nauk. S.S.S.R., Dpklady, 59 (3), PP. 545-548. (1948B), Akad, Nauk. S.S.S.R., Doklady, 59 (4), PP. 739-742. Service, H. (1943), Gold Coast Geol. Surv., Mem. No. 5. Shannon, E. V. (1920), Am. Mineral., 5, pp. 84-860 Shibata, H. (1952A), Tokyo Bundrika Daigaku, (Tokyo Univ. Lit. Sci.) Geol. Min. Inst., Science Reports, Sec. C, 2, No. 10-11. (1952B), Tokyo Bundrika Daigaku (Tokyo Univ. Lit. Sci.) Geol. Min. Inst., Science Reports, Sec. C, 2, No. 12. Simpson, E. S. (1912), Geol. Survey West. Australia, Bull. 48, pp. 95-99. (1927), Journ. Roy. Soc. West. Australia, 13, pp. 37-49. (1932), Journ. Roy. Soc. West. Australia, 18, pp. 61-74. Sollas, W. J. (1889), Proc. Roy Irish Acad., Sec. 3, I, pp. 379-380. Spence, H. S. (1930), Am. Mineral., 15, p. 438. Starkl, G. (1883), Jahrb. der K.-K. Geol. Reichsanstalt, Vienna, 33, pp. 653-656. Sterrett, D. B. (1923), U.S.G.S. Bull., 740, pp. 12, 65, 133. Stevens, R. E. (1938), Am. Mineral., 23, pp. 607-628. 239

Stevens, E. E. (1946), U.S.G.S. Bull., 950, pp. 101-119. Stoll, W. C. (1950), U.S.G.S., Prof. Paper 229, p. 62. Sudo, T. (1949), Bull. Chem. Soc. Japan, 22, pp. 25-33. Sugi, K. (1940), Mem. Fac. Sci., Kyusyu Imp. Univ., Ser. D, (1), p. 7. Svanberg, L. F. (1844), Konglig. Veten. Akad. Stock. Ofver., 1, p. 219. Taddei, C. (1940)9 Schweiz Min. Petrog. Mitt., 20, pp. 247-252. Tchirvinskii, P. N. (1948), Mem. Soc. Russe Min., 77 (3), pp. 246-9, Thiebaut, J. L. (1925), Contribution a l'etude des Sediments Argilocalcaires du Bassin de Paris: Nancy. Cit. in Min. Mag., 24, p. 621. Thomson, T. (1828), Annals of the Lyceum of Natural History of New York, 3, p. 47. ~- - (1836), Outlines of Mineralogy, Geology, and Mineral Analysis: 2 volumes, London. Tolansky, S. (1948), Multiple-Beam Interferometry of Surfaces and Films: The Clarendon Press, Oxford. Tschermak, G. (1877), Zeit. Krist., 2, pp. 14-50. (1878), Zeit. Krist., 3, pp. 122-167. Turin, B. A. (1944), Akad. Nauk. U.S.SoR. Izv., Ser. Geol. 2. vanBiljon, S. (1940), Trans. Geol. Soc. South Africa, 42, pp. 123-219. Venkayya, E. (1949), Proc. Indian Acad. Science, 30, Sec. A, pp. 74-77. Volk, G. W. (1939), Am. Mineral., 24, pp. 255-266. Wagner, P. A. (1921), Trans. Geol. Soc. S. Africa, 24, p. 73. (1928), Geol. Surv. Union of South Africa, Mem. 26. Wahl, W. (1925), Fennia, 45 (No. 20), 127 pp. Walker, T. L., and Parsons, A. L. (1924), Univ. of Toronto Studies, Geol. Series, No. 17, pp. 46-50. (1926), Univ. of Toronto Studies, Geol. Series, No. 22, pp. 5-14. Webb, R. W. (1939), Am. Mineral., 24, pp. 123-129. Weinschenk, E. (1896), Zeit. Krist., 26, p. 471. 240

Wells, Ro C., and Brannock, W. W. (1946), U.S.G.S. Bull., 950, pp. 121-127. Wherry, E. T. (1915), Proc. U. S. Nat. Mus., 49, pp. 463-467. Whitmore, D. R. E., Berry, L. G., and Hawley, J. E. (1946), Am. Mineral., 31, pp. 1-21. Wijkerslooth, P. (1943), Maden Tetkik ve Arama Eustitusu Mecmuasi, Ankara, 8, pp. 91-101. Winchell, A. N. (1925), Amer. Journ. Sci., 209, pp. 309-327; 415-430. (1927), Am. Mineral., 12, pp. 267-279. (1932), Am. Mineral., 17, pp. 551-553. (1935), Am. Mineral., 20, pp. 773-779. (1942), Am. Mineral., 27, pp. 114-130. and Winchell, H. (1951), Elements of Optical Mineralogy: Part II, New York. Wolff, J. E., and Palache, C. (1902), Zeit. Krist., 36, pp. 438-448. Wood, W. A. (1937), Proc. Roy. Soco, Ser. A, 163, pp. 199-204. Woodward, H. H. (1951), Am. Mineral., 36, pp. 869-883. Wright, F. E. (1914), Amer. Journ. Sci., 38, pp. 305-308. Wulfing, E. A. (1886), Ber. Deutschen Chem. Gesell., 19, pp. 24332438. Yamada, H., and Sugiura, K. (1950), Journ. Jap. Assoc. Mineralogists, Petrologists and Econ, Geologists, 34, pp. 122-128. Yoshimura, T. (1939), Journ. Fac. Sci., Hokkaido Imp. Univ., Ser. 4, Vol. 4, No. 3-4, pp. 313-452. Zambonini, F. (1919), Mem. Descr Carta Geol. Italia, 7, Part 2, p. 124. (1922), Rend. Accad. naz. Lincei, Ser0 5, 31 (1), pp. 147-151. Zealley, A. E. Vo (1917), Proc. Rhodesia Sci. Assoc., 16, pp. 17-23. 241