X HW*' 2*^**' IV^ \1^; S-^f^^l * IP Linna'ite 1 86 Sychnodymite _ 187 Erythrite or cobalt bloom 187 Asbolite 1 87 4. Arsenopyrite; mispickel or arsenical pyrites 1 89 5. Lollingite; leucopyrite 18!) 6. Pyrites 190 7. Molybdenite 193 III. Halides: 1. Halite; sodium chloride; or common salt 195 2. Fluorite 213 3. Cryolite 214 IV. Oxides: 1. Silica 215 Quartz 215 Flint 216 Buhrstone 217 Tripoli 217 Diatomaceous or infusorial earth 218 2. Corundum and emery 220 3. Bauxite 229 4. Diaspore 239 5. Gibbsite; hydrargillite 239 6. Ocher 239 7. Ilmenite; menaccanite; or titanic iron 245 8. Rutile 245 9. Chromite 246 10. Manganese oxides 252 Franklinite 253 Hausmannite 253 159 IQQ KEPOBT OF NATIONAL MUSEUM, 1899. IV. Oxides Continued. 10. Manganese oxides Continued. Braunite 254 Polianite /a * Pyrolusite Manganite Psilomelane Wad or bog manganese 255 V. Carbonates: 1. Calcium carbonate Calcite; calc spar; Iceland *par 258 Chalk Limestones; mortars; and cements Portland cement Roman cement Playing marbles 270 Lithographic limestones 270 2. Dolomite 274 3. Magnesite 275 4. Witherite 279 5. Strontianite 279 6. Rhodochrosite; dialogite 280 7. Natron, the nitrum of the ancients 280 8. Trona; urao 281 VI. Silicates: 1. Feldspars 281 2. Micas 283 3. Asbestos 2% 4. Garnet 307 5. Zircon 308 6. Spodumene and petalite 308 7. Lazurite; lapis lazuli; or native ultramarine 309 8. Allanite; orthite 311 9. Gadolinite 313 10. Cerite 314 11. Rhodonite 314 12. Steatite; talc; and soapstone 315 13. Pyrophyllite; agalmatolite; and pagodite 322 14. Sepiolite; meerschaum 323 15. Clays 325 VII. Niobates and tantalates: 1. Columbite and tantalite 353 2. Yttrotantalite 354 3. Samarskite 354 4. Wolframite and Hiibnerite 355 5. Scheelite 355 VIIL Phosphates: 1. Apatite; rock phosphates; guano, etc 356 2. Monazite 3g3 3. Vanadinite 3g7 4. Descloizite 388 5. Amblygonite gcjQ 6. Triphylite and iithiophilite 39] CONTENTS. 161 IX. Nitrates: Page. 1. Niter, potassium nitrate 391 2. Soda niter 392 3. Nitro-calcite 394 X. Borates: 1. Borax or tincal; borate of soda 396 2. Ulexite; boronatrocalcite 397 3. Colemanite 397 4. Boracite or stassfurtite; borate of magnesia 397 XI. Uranates: 1. Uraninite; pitchblende 402 XII. Sulphates: NJB. Barite; heavy spar 405 2. Gypsum 406 3. Celestite 411 4. Mirabilite; or Glauber salt 412 5. Glauberite 415 6. Thenardite 415 7. Epsomite; Epsom salts 415 8. Polyhalite 416 9. Kainite 416 10. Kieserite 416 11. Alums: Kalinite 416 Tschermigite 416 Aluminite 419 Alunite 419 Alum slate or shale 421 XIII. Hydrocarbon compounds: 1. Coal series 423 Peat 424 Lignite or brown coal 425 Bituminous coals 426 Anthracite coal 427 2. Bitumen series 429 Marsh gas; natural gas 433 Petroleum 434 Asphaitum; mineral pitch 441 Manjak 445 Elaterite; mineral caoutchouc 446 Wurtzillite '. 446 Albertite 446 Grahamite 447 Carbonite or natural coke 449 Uintaite; gilsonite 450 3. Ozokerite; mineral wax; native paraffin 451 4. Resins 455 Succinite; amber 455 Retinite 456 Chemawinite 456 Gum copal 457 NAT MU8 99 11 162 REPORT OF NATIONAL MUSEUM, 1899. XIV. Miscellaneous: Page. 1. Grindstones; whetstones; and hones 463 2. Pumice 470 3. Rottenstone 473 4. Madstones / 474 5. Molding sand 474 6. Mineral waters 477 7. Road-making materials 482 LIST OF ILLUSTRATIONS. PLATES. Facing page 1. View showing wall and rail cases and installation of nonmetallic minerals on gallery of southwest court, U. S. National Museum. Looking west 155 2. View showing rail case and installation of nonmetallic minerals in gallery of southwest court of U. S. National Museum. Looking north 1 55 3. Views in graphite mine near Hague, Warren County, New York. From photographs by C. D. Walcott 170 4. Section of the salt deposits at Stassfurt. From the Transactions of the Edinburgh Geological Society, V, 1884, p. 11 1 204 5. Views of brine-evaporating tanks at Syracuse, New York. From photo- graphs by I. P. Bishop 210 6. View of Tripoli mines in Carthage, Missouri 218 7. Deposit of diatomaceous earth, Great Bend of Pitt River, Shasta County, California. From a photograph by J. S. Diller 219 8. Map showing distribution of corundum and peridotite in the eastern United States. After J. V. Lewis, Bulletin II, North Carolina Geo- logical Survey 222 9. Microstructure of emery. After Tschermak, Mineralogische und Petro- graphische Mittheilungen XIV, Part 4 224 10. Section showing the formation of manganese deposits from dec-ay of limestone. After Penrose, Annual Report Geological Survey of Arkan- sas, I, 1890 252 11. Botryoidal psilomelane, Crimora, Virginia. Specimen No. 66722, U.S.N.M 255 12. Views showing occurrence Calcite in Iceland. After Thoroddsen 259 13. View in a cement quarry near Whitehall, Ulster County, New York. From a photograph by N. H. Barton 268 14. View in a soapstone quarry, Lafayette, Pennsylvania 319 15. Microsections showing the appearance of (1) kaolinite and (2) washed kaolin 330 16. Microsections showing the appearance of (1) halloysite and (2) ledaclay. 331 17. Microsections showing the appearance of (1) Albany County, Wyoming, clay and (2) fuller's earth 332 18. Leda clays, Lewiston, Maine. From a photograph by L. H. Merrill 333 19. View in a Kaolin pit, Delaware County, Pennsylvania 339 20. Map showing phosphate regions of Florida. After G. H. Eldridge 366 21. Borax mine near Daggett, California. Interior and exterior views 398 22. View of a gypsum quarry. From a photograph by the Iowa Geological Survey 408 23. Peat beds overlying gold-bearing gravels, Mias, Russia. From a photo- graph by A. M. Miller , , 424 163 164 REPORT OF NATIONAL MUSEUM, 1899. Facing page, 24. Map showing the developed coal fields of the United States. From the Eeport of the Eleventh Census 25 Map showing areas where bitumen occurs in the United States and Canada. From the Report of the Tenth Census . . 26 Plan of Pitch Lake, Trinidad. After S. F. Peckham 442 27. Nodule of gum copal from Congo River region, Africa. Specimen No. 62717, U.S.N.M 457 28. Microsection of mica schist used in making whetstone. Fig. 1, cut across foliation. Fig. 2, cut parallel to foliation 467 29. View of Novaculite Quarry, Arkansas. After Griswold, Annual Report of the Geological Survey of Arkansas, III, 1890 468 30. Microsections showing the appearance of (1) Arkansas Novaculite and (2) Ratisbon razor hone. The dark bodies in (2) are garnets 470 TEXT FIGURES. Page. 1. Block of limestone with alternating bands of sulphur. Sicily, Italy. Spec- imen No. 60932, U.S.N.M .--- 179 2. Cluster of halite crystals, Stassfurt, Germany. Specimen No. 40222, U. S. N. M 195 3. Geological section of Petite Anse Island, Louisiana 201 4. Cluster of sylvite crystals, Stassfurt, Germany. Specimen No. 40223, U.S.N.M 203 5. Pisolitic bauxite. Bartow County, Georgia. Specimen No. 63335, U.S.N.M. 229 6. Map showing geological relations of Georgia and Alabama bauxite deposits. After C. W. Hayes 235 7. Section showing relation of bauxite to mantle of residual clay in Georgia. After C. W. Hayes 236 8. Section across paint mine at Lehigh Gap, Pennsylvania. After C. E. Hesse . 242 9. Section of mica veins in Yancey County, North Carolina. After W. C. Kerr 288 10. Asbestos fibers. After G. P. Merrill, Proceedings of the U. S. National Museum, XVIII, p. 283 297 11. Serpentine asbestos in massive serpentine. Specimen No. 72836 302 12. Map of Nitrate region, Chile. After Fuchs and De Launay 393 13. Section through Sulphur Mountain, California. After S. F. Peckham . . 432 GUIDE TO THE STUDY OF THE COLLECTIONS IN THE SECTION OF APPLIED GEOLOGY. THE NONMETALLIC MINERALS. By GEORGE P. MERRILL, Curator, Division of Physical and Chemical Geology and Head Curator of the Department. I. ELEMENTS. 1. CARBON. The numerous compounds of which carbon forms the chief constit- uent are widely variable in their physical properties and origin. As occurring in nature few of its members possess a definite chemical composition such as would constitute a true mineral species, and they must for the most part be looked upon as indefinite admixtures in which carbon, hydrogen, and oxygen play the more important roles. For present purposes the entire group may be best considered under the heads of (1) The Pure Carbon series; (2) The Coal series, and (3) The Bitumen series, the distinctions being based mainly on the gradu- ally increasing amounts of volatile hydrocarbons, a change which is accompanied by a variation in physical condition from the hardest of known substances through plastic and liquid to gaseous forms. Here will be considered only the members of the pure carbon series, the others being discussed under the head of hydrocarbon compounds. DIAMOND. This mineral crystallizes in the isometric system, with a tendency toward octahedral forms, the crystals showing curved and striated surfaces. (Specimen No. 53558, U.S.N.M.) The hardness is great, 10 of Dana's scale; the specific gravity varies from 3.1 in the carbonados to 3.5 in good clear crystals. The luster is adamantine; the colors, white or colorless, through yellow, red, orange, green, brown to black. The transparent and highly refractive forms are of value as gems, and can best be discussed in works upon this subject. We have to do here rather with the rough, confused crystalline aggre- gates or rounded forms, translucent to opaque, which, though of no value as gems, are of the greatest utility in the arts. To such 165 166 REPORT OF NATIONAL MUSEUM, 1899. forms the name Hack diamond, bort, and carbonado are applied. (Speci- mens Nos. 53668-53671, U.S.N.M.) Origin and Occurrence. The origin of the diamond has long been a matter of discussion. A small proportion of the diamonds of the world are found in alluvial deposits of gravel or sand. In the South African fields they occur in a so-called blue gravel, formed, according to Lewis, along the line of contact between an eruptive rock (perido- tite) and highly carbonaceous shales. They were regarded by Lewis as originating through the crystallization of the carbon of the shales by the heat of the molten rock. De Launay states, however, that there is no necessary connection betw r een the shales and the diamond, and shows with apparent conclusiveness that the latter occur often in a broken and fragmental condition, such as to indicate beyond doubt that they originated at greater depths and were brought upward as phenocrysts in the molten magma at the time of its intrusion. The primary origin of the diamonds he regards as through the crystalliza- tion, under great pressure, of the carbon contained in the basic magma in the form of metallic carbides. The diamond-bearing rock as above noted is a peridotite often brec- ciated and more or less serpentinized (Specimen No. 62108, U.S.N.M.). The blue and green gravel formed by the decomposition of this rock is shown in Specimen No. 73188, U.S.N.M. With these are others of the associated, eruptive, and metamorphic rocks, as melaphyr (Specimen No. 73184, U.S.N.M.), quartzite (Specimen No. 73185, U.S.N.M.), shale (Specimen No. 73186, U.S.N.M.), and basalt (Specimen No. 73187, U.S.N.M). Whether or not a similar origin to that outlined above can be attrib- uted to the Brazilian diamonds is as yet unproven. Their occurrence and association with detrital materials resulting from the breaking down of older rocks, with which they may or may not have been originally associated, renders the problem obscure and difficult of solution. According to Kunz, 1 95 per cent of all diamonds at present obtained come from the Kimberly Mines, Griqua Land, west South Africa; of these, some 47 per cent are bort. The remainder come from Brazil, India, and Borneo. A few have been found in North America, the Ural Mountains, and New South Wales, but these countries are not recognized as regular and constant sources of supply. Uses. The material, aside from its use as a gem, owes its chief value to its great hardness, and is used as an abrading and cutting medium in cutting diamonds and other gems, glass, and hard materials in gen- eral, such as can not be worked by softer and cheaper substances. With the introduction of machinery into mining and quarrying there 1 Gems and Precious Stones. New York, 1890. THE NONMETALLIC MINERALS. 167 has arisen a constant and growing demand for black diamonds, or bort, for the cutting edges of diamond drills, and to a less extent for teeth to diamond saws. (Specimens Nos. 53668 to 53670, U.S.N.M.) According to a writer in the Iron Age 1 the crystallized diamond is not suitable for these purposes owi ng to its cleavage property. The best bort or "carbonado" comes, it is said, from Bahia, Brazil, where it is found as small, black pebbles in river gravels. The ordinary sizes used for drills weigh but from one- half to 1 carat, but in special cases pieces weighing from 4 to 6 carats are used. It is stated that the crowns of large drills, 10 inches in diameter, armed with the best grade of carbonado, are sometimes valued as high as $10,000. BIBLIOGRAPHY. M. BABINET. The Diamond and other precious stones. Report of the Smithsonian Institution, 1870, p. 333. A DAUBREE. Annales des Mines, 7th ser., IX, 1876, p. 130. Remarking on the occurrence of platinum associated with peridotites, he calls attention to the fact that Maskelyne had shown the diamonds of South Africa and Borneo to occur in a decomposed peridotite. ORVILLE A. DERBY. Geology of the Diamantiferous Region of the Province of Parand, Brazil. American Journal of Science, XVIII, 1879, p. 310. Geology of the Diamond. American Journal of Science, XXIII, 1882, p. 97. H. COHEN. Igneous origin of the Diamond. Proceedings, Manchester Literary and Philosophical Society, 1884, p. 5. H. CARVILL LEWIS. The Genesis of the Diamond. Science, VIII, 1886, p. 345. GARDNER F. WILLIAMS. The Diamond Mines of South Africa. Transactions of the American Institute of Mining Engineers, XV, 1886, p. 392. ORVILLE A. DERBY. The Genesis of the Diamond. Science, IX, 1887, p. 57. Discovery of Diamonds in a Meteoric Stone. Nature, XXXVII, 1887, p. 110. Diamond Mining in Ceylon. Engineering and Mining Journal, XLIX, 1890, p. 678. A. MERVYN SMITH. The Diamond Fields of India. Engineering and Mining Journal, LIII, 1892, p. 454. OLIVER WHIFFLE HUNTINGTON. Diamonds in Meteorites. Science, XX, 1892, p. 15. Diamonds in Meteoric Stones. The American Geologist, XI, 1893, p. 282. (Abstract of paper by H. Moissan, Comptes Rendus 1893, pp. 116 and 228.) HENRI MOISSAN. Study of the Diamantiferous Sands of Brazil. Engineering and Mining Journal, LXII, 1896, p. 222. HENRY CARVILL LEWIS. I. Papers and Notes on the Genesis and Matrix of the Diamond, edited by Prof. T. G. Bonney. The Geological Magazine, IV, 1897, p. 366. Sir WILLIAM CROOKES. Diamonds. Nature, LV, 1897, p. 325. 1 Volume XXXVI, December 24, 1885, p. 11. 168 REPORT OF NATIONAL MUSEUM, 1899. L. DE LA UNA Y. Les Diamants du Cap. ORVILLB A. 'DERBY. Brazilian Evidence on the Genesis of the Diamond. The Journal of Geology, VI, 1898, p. 121. H. W. FCRMISS. Carbons in Brazil. U. S. Consular Reports, 1898, p. 604. See also Engineering and Mining Journal, LXVI, 1898, p. 608. M. J. KLINCKB. Gites Diamantiferes de la Republique sud-Africaine. Annales des Mines, XIV, 1898, p. 563. GRAPHITE. Graphite, plumbago, or black lead, as it is variously called, is a dark steel gray to black lustrous mineral with a black streak; hardness of but 1.2, and a specific gravity of from 2.25 to 2.27. The prevailing form of the mineral is scaly or broadly foliated (Speci- men No. 51007, U. S. N. M.), with a bright luster, but it is sometimes quite massive (Specimen No. 61138, U. S. N. M.) and columnar (Speci- men No. 59976, U.S.N.M.) or earthy, with a dull coal-like luster (Specimens Nos. 64795 and 63133, U.S.N.M.). Its most characteristic features are its softness, greasy feeling, and property of soiling everything with which it conies in contact. Molybdenite, the sulphide of molybdenum, is the only mineral with which it is likely to become confounded. This last, however, though very similar in general appearance, gives a streak with a slight green- ish tinge, and when fused with soda before the blowpipe yields a sul- phur reaction. Chemically, graphite is nearly pure carbon. The name black lead is therefore erroneous and misleading, but has become too firmly established to be easily eradicated. The analyses given below show the composition of some of the purest natural graphites. Locality. Carbon. Ash. Volatile matter. Ceylon 98. 817 280 90 Buckingham, Canada 97. 626 Do 99815 1.78 076 .594 109 As mined the material is almost invariably contaminated by mechan- ically admixed impurities. Thus the Canadian material (Specimens Nos. 59977, 62153, U.S.N.M.) as mined yields from 22.38 to 30.51 per cent of graphite; the best Bavarian, 53.80 per cent (Specimen No. 52050, U.S.N.M.). The grade of ore that can be economically worked naturally depends upon the character of the impurities and the extent and accessibility of the deposit. It is said 1 that deposits at Ticonde- roga, New York, have been worked in which there was but 6 per cent of graphite (Specimen No. 37825, U. S. N. M.). Occurrence and origin, Graphite occurs mainly in the older crystal- 1 Engineering and Mining Journal, LXV, 1898, p. 256. THE NONMETALLIC MINERALS. 169 line metamorphic rocks, both siliceous and calcareous, sometimes in the form of disseminated scales, as in the crystalline limestone of Essex County, New York (Specimen No. 37825, U.S.N.M), or in embedded masses, streaks, and lumps, often of such dimensions that single blocks of several hundred pounds weight are obtainable. (Specimen No. 59976, U.S.N.M.) It is also found in the form of veins. The fact that the mineral is carbon, one of the constituents of animal and vegetable life, has led many authorities to regard it, like coal, as of vegetable origin. While this view is very plausible it can not, how- ever, be regarded as in all cases proven. That graphite may be formed independently of organic life is shown by its presence in cast iron, where it has crystalized out, on cooling, in the form of bright metallic scales. See Specimens Nos. 51298 and 51312 in the metallurgical series of the manufacture of iron. Carbon is also found in meteorites which are plainly of igneous origin, and which have thus far yielded no certain traces of either plant or animal organisms. It is, however, a well-known fact that coal itself of organic origin has in some cases been converted into graphite through metamorphic agencies, and intermediate stages like the graphitic anthracite of Newport, Rhode Island, afford good illus- trations of such transitions. (Specimen No. 59099, U.S.N.M.) Certain European authorities 1 have shown that amorphous carbonaceous par- ticles in clay slates have been converted into graphite by the metamor- phosing influence of intruded igneous rocks. Prof. J. S. Newberry described an occurrence of this nature in the coal fields of Sonora, Mexico. 2 He says: All the western portion of this coal field seems to be much broken by trap dikes which have everywhere metamorphosed the coal and converted it into anthracite. At the locality examined the metamorphic action has been extreme, converting most of the coal into a brilliant but somewhat friable anthracite, containing 3 or 4 per cent of volatile matter. At an outcrop of one of the beds, however, the coal was found converted into graphite, which has a laminated structure, but is unctuous to the touch and marks paper like a lead pencil. The metamorphism is much more complete than at Newport (Rhode Island) [Specimen No. 59099, U.S.N.M.], furnishing the best example yet known to me of the conversion of a bed of coal into graphite. In New York State, and in Canada, graphite occurs in Laurentian rocks, both in beds and in veins, a portion of the latter being appar- ently true fissure veins and others shrinkage cracks or segregation veins which traverse in countless numbers-the containing rocks. It is said 3 that in the Canadian regions (Specimens Nos. 51007, 59976, U.S.N.M.), the deposits occur generally in limestone or in their immediate vicinity, and that granular varieties of the rock often contain large crystalline 1 Beck and Luzi, Berichte der Deutschen Chemischen Gesellschaft, 1891, p. 24. 2 Schoolof Mines Quarterly, VIII, 1887, p. 334. 3 See On the Graphite of the Laurentian of Canada, by J. W. Dawson, Proceedings of the Geological Society of London, XXV, 1870, p. 112, and an article on Graphite by Prof. J. F. Kemp in The Mineral Industry, II, 1893, p. 335. 170 REPORT OF NATIONAL MUSEUM, 1899. plates of plumbago. At other times the mineral is so finely dissemi- nated as to give a bluish-gray color to the limestone, and the distribu- tion of the bands thus colored seems to mark the stratification ot the rock. Further, the plumbago is not confined to the limestones; large crystalline scales of it are occasionally disseminated in pyroxene rock or pyrallolite, and sometimes in quartzite and in feldspathic rocks, or even in magnetic oxide of iron. In addition to these bedded forms, there are also true veins in which graphite occurs associated with cal- cite, quartz, orthoclase, or pyroxene, and either in disseminated scales, in detached masses, or in bands or layers separated from each other and from the wall rock "by feldspar, pyroxene, and quartz. Kemp describes * the graphite deposit near Ticonderoga, New York (Specimens Nos. 37825, 66759, U.S.N.M.), as in the form of a true fissure vein, cutting the lamination of the gneissic walls at nearly right angles. The wall rock is a garnetiferous gneiss, with an east and west strike, and the vein runs at the u big mine" north 12 west, and dips 55 west. The vein filling, he says, was evidently orthoclase (or microcline) with quartz and biotite and pockets of calcite. Besides graphite, it con- tained tourmaline, apatite, pyrite, and sphene. Walcott 8 describes the graphite at the mines 4 miles west of Haguo, on Lake George, New York, as occurring in Algonkian rocks, and as probably of organic origin. At the mines the alternating layers of graphite shale or schist form a bed varying from 3 to 13 feet in thickness. The outcrop may be traced for a mile or more. The garnetiferous sandstones form a strong ledge above and below the graphite bed. The appearance is that of a fossil coal bed, the alteration having changed the coal to graphite and the sandstone to indurated, garnetiferous, almost quartzitic sandstones. The character of the graphite bed is well shown in the accompanying plate, from a photograph tat en by me in 1890. It is here a little over 9 feet in thickness and is formed of alternating layers of highly graphitic sandy shale and schist. [See Plate 3.] According to J. Walther 3 the Ceylonese graphite (Specimens Nos. 66857, 62073, U.S.N.M.) occurs in coarsely foliated or stalky masses in veins in gneiss which, where mined, is decomposed to the condition of laterite. The veins are regarded as true fissures, and vary from 12 to 22 cm. (about 4f to 8f inches) in width. The graphite of Northern Moravia occurs in gray to black crystal- line granular Archaean limestone interbedded with amphibolitos juid muscovite gneiss, the limestone itself being often serpentinous, in this respect apparently resembling the graphitic portions of the ophical- cites of Essex County, New York. (Specimen No. 70084, U.S.N.M.). The material is quite impure, showing on the average but 53 per cent of carbon and 44 per cent of ash, the latter being made up largely of 1 Preliminary Report on the Geology of Essex County, Contributions from the Geo- logical Department of Columbia College, 1893, pp. 452, 453. "Bulletin of the Geological Society of America, X, 1898, p. 227. 3 Records of the Geological Survey of India, XXIV, 1891, p. 42. Report of U. S. National Museum, 1 899. Mer PLATE 3. VIEWS IN GRAPHITE MINE NEAR HAGUE, WARREN COUNTY, NEW YORK. From photographs by Charles D. \Valcott. THE NONMETALLIC MINERALS. 171 silica and iron oxide, with a little sulphur, magnesia, and alumina. This graphite is regarded as originating through the metamorphism of vegetable matter included in the original sediments, the agencies of metamorphism being both igneous intrusions and the heat and pressure incidental to the folding of the beds. 1 As to so much of the graphite as occurs in beds there seems, then, little doubt as to its origin from plant remains which may be imagined to have existed in the form of seaweeds or to have been derived from diffused bituminous matter. The origin of the vein material is not so evident, though it seems probable that it is due to the metamorphism of bituminous matter segregated into veins, like those of albertite in New Brunswick or of gilsonite, etc., in Utah. Kemp states that the Ticonderoga graphite must have reached the fissure as some volatile or liquid hydrocarbon, such as petroleum, and become metamorphosed in time to its present state. Walther believes the Ceylon material to have originated by the reduction of carburetted vapors. (See also under origin of diamonds, p. 166.) The total quantity of carbon in the form of graphite in the Lauren- tian rocks of Canada has been estimated by Dawson as equal to that in any similar areas of the Carboniferous system of Pennsylvania. Sources. The chief sources of the graphite of commerce are Austria and Ceylon. Other sources of commercial importance are Germany, Italy, Siberia (Specimen No. 61138, U.S.N.M.), the United States, and Canada. The chief deposits of commercial value in the United States are at Ticonderoga, New York, where the graphite occurs in a granu- lar quartz rock, or, according to J. F. Kemp, in "Elliptical Chimneys in Gneiss which are filled with Calcite and Graphite." An earth}', impure graphite, said to be suitable for foundry facings, is mined near Newport, Rhode Island (Specimen No. 53797, U.S. X.M.). About one hundred years ago the material was mined in Bucks County, Pennsyl- vania. Other American localities represented in the collections are Bloomingdale, New Jersey (Specimen No. 56272, U.S.N.M.); Clinton- ville, New York (Specimen No. 31597, U.S.N.M.); Hague, Warren County, New York (Specimen No. 63132, U.S.N.M.); Raleigh, Wake County, North Carolina (Specimen No. 63133. U.S.N.M.); Lehigh and Berks counties, Pennsylvania (Specimens Nos. 66952; 66953, U.S. N. M. ) ; Salt Sulphur Springs, West Virginia (Specimen No. 63423, U.S.N.M.); St. Johns, Tooele County, Utah (Specimen No. 62721, U.S.N.M.). ' Graphite is a very common mineral in the Laurentian rocks of Canada. The most important known localities are north of the Ottawa River, in the townships of Buckingham, Lochaber, and Grenville (Specimens Nos. 59976,. 51007, U.S.N.M.). At Buckingham it is stated masses of graphite have been obtained weighing nearly 5,000 pounds. JJahrbuch k. k. Geologische Reichsanstalt, 1897, XL VII, p. 21. 172 REPORT OF NATIONAL MUSEUM, 1899. At Grenville the graphite occurs in a gangue consisting mainly of pyroxene, wollastonite, feldspar, and quartz, while the country rock is limestone. Blocks of graphite have been obtained weighing from 700 to 1,500 pounds. 1 Graphite is also found in Japan (Specimen No. 34359, U.S.N.M.), Australia (Specimen No. 62177, U.S.N.M.), New Zealand (Specimens Nos 17796 and 64795, U.S.N.M.), Greenland (Specimen No. 65374, U.S.N.M.), Guatemala (Specimen No. 33990, U.S.N.M.), Germany, and in almost all the Austrian provinces, the most important and best known deposits being those of Kaiserberg at St. Michel, where there are five parallel beds occurring in a grayish black graphite schist, the beds varying from a few inches to 6 yards. The only workable deposit in Germany is stated to be at Passau in Bavaria. The material occurs in a feldspathic gneiss, seeming to take the place of the mica (Specimen No. 52050, U.S.N.M.). The beds have been worked chiefly by peasants for centuries, and the output used mainly for crucibles. 2 " lf ses , Graphite is used in the manufacture of "lead" pencils, lubricants, stove blacking, paints, refractory crucibles, and for foun- dry facings. In the manufacture of pencils only the purest and best varieties are used, and high grades only can be utilized for lubricants (Specimens Nos. 51608-51619, U.S.N.M.). For the other purposes mentioned impure materials can be made to answer. In the manufac- ture of the Dixon crucibles (Specimens Nos. 51598-51600, U.S.N.M.) a mixture of 50 per cent graphite, 33 per cent of clay, and 17 per cent of sand is used. Preparation. In nature graphite is usually associated with harder and heavier materials, which it is necessary to get rid of before the material is of value. In New York it is the custom to crush the rock in a battery of stamps, such as are used in gold mining, and then separate the graphite by washing, its lighter specific gravity permit- ting it to be floated off on water, while the heavy, injurious constitu- ents are left behind. Mica, owing to its scaly form, can not be separated in this manner, and hence micaceous ores of the mineral are of little if any value. An improvement in the manufacture of plumbago or graphite has been described in a recent patent specification. Graphite, crushed and passed through a sieve of from 120 to 150 meshes per inch, is stirred into a saturated solution of alum or aluminum sulphate at a temperature of 212 F. ; steatite is then added, and more water, if required. After mixing, excess of water is evaporated until a consistency suited to grinding in a chilled steel or other mixer is obtained. More graphite may here be added; then, after thorough grinding, the material may be compressed into cakes for household use, or is ready for the manu- 1 Descriptive Catalogue of Economic Minerals of Canada, 1876, p. 122. 2 The Journal of the Iron and Steel Institute, 1890, p. 739. THE NONMETALLIC MINERALS. 173 fact lire of pencils or crucibles. The average formula of the mixture is: Graphite, 80 parts; steatite, soapstone, or talc, 14 parts; alum, 6 parts; but this varies with the purpose to which the material is to be applied. When several different kinds of graphite have to be employed, the richest in carbon is first mixed into the alum solution. By this process graphites previously regarded as incapable of being compacted are utilized, and are improved in polishing power. For pencils the material may be hard without being brittle, and black without being soft, while crucibles made from the treated graphite are at once harder, more durable, and lighter. 1 Prices. The value of the mineral varies with its quality. In 1899 the crude lump was reported as worth $8 a ton and the pulverized $30. The annual output as given 2 for the principal countries is as follows: World 's production of graphite. Year. Austria. Canada. Ceylon. Germany. India. Italy. United States. 1892 Metric tons. 20, 978 Metric tons. 151 Metric tons. 21,300 Metric tons. 4,036 Metric tons. (a) Metric tons. 1 , 645 Metric ton*. 707 1893 23 807 Nil 21 900 3 140 (a) 1 465 634 1894 1895 24,121 28 443 C3 199 10,718 13 711 3, 133 3 751 1,623 1,575 349 171 18% 1897 1898 35,972 38,504 33 062 126 396 1 107 10,463 619,275 678 509 5,248 3,861 4 593 () 61 2 3, 148 5, 650 6 435 184 450 ,X'>4 a Not reported in the Government statistics. BIBLIOGRAPHY. b Exports. J. W. DAWSON. On the Graphite of the Laurentian of Canada. Quarterly Journal Geological Society of London, XXVI, 1870, p. 112. M. BONNEFOY. M^moire sur la Geologie et 1' Exploitation des Gites de Graphite de la Boheme Meridionale. Annales des Mines. 7th Ser., XV. 1879, p. 157. JOHN S. NEWBERRY. The Origin of Graphite. School of Mines Quarterly, VIII, 1887, p. 334. Der Graphitbergbau auf Ceylon. Berg- und Huttenmannische Zeitung, XLVII, 1888, p. 322. J. WALTHER. Ueber Graphitgange in zersetztem Gneiss (Laterit) von Ceylon. Zeitschrift der Deutschen Geologischen Gesellschaft, XLI, 1889, p. 359. A. PALLAUSCH. Die Graphitbergbaue im siidlichen Bohmen. Berg- und Huttenmiinnisches Jahrbuch, XXXVII, p. 95, 1889. T. ANDREE. Graphite Mining in Austria and Bavaria. (Abstract.) . Journal of the Iron and Steel Institute, 1890, p. 738. 1 Engineering and Mining Journal, LVI1I, 1894, p. 440. 3 The Mineral Industry, VI, 1897; VIII, 1899. 174 REPORT OF NATIONAL MUSEUM, 1899. J. POSTLETHWAITE. The Borrowdale Plumbago; its Mode of Occurrence and Probable of the Geological Society of London, Session, 1889-1890, p. 124. ischen Gesellschaft, XXIV, pp. 4085-4095. 1891.) Neues Jahrbuch fur Mineralogie, Geologic und Paleontologie. 1393. II, P< 2 p. 241. (Abstract.) E.WEINSCHENK. Zur Kenntniss der Graphitlagerstatten. Chemisch-geologisc Studien von Dr. Ernst Weinschenk. 1 Die Graphitlagerstatten des bayerischen Grenzgebirges. Habihtations- schrift zur Erlangung der venia legendi an der K. technischen Hochschule. Miinchen, 1897. FRANZ KRETSCHMER. The Graphite Deposits of Northern Moravia. Transactions of the North of England Institute of Mining. and Mechanical Engineer, XLVII, 1898, p. 87. 2. SULPHUR. Color of the mineral when pure yellow, sometimes brownish, red- dish, or gray through impurities. Hardness, 1.5 to 2.5. Specific gravity, 2.05. Insoluble in water or acids. Luster resinous. Occurs native in beautiful crystals (Specimens Nos. 53115, 53116, and 60660, U.S.N.M.) or in massive (Specimens Nos. 16092, 60849, U.S.N.M.), stalactitic and spheriodal forms (Specimens Nos. 57137 and 60864, U.S.N.M.). Once seen the mineral is as a rule readily recognized, and all possible doubts are set at rest by its ready inflammability, burning with a faint bluish flame and giving the irritating odors of suiphurous anhydride. In nature often impure through the presence of clay and bituminous matters; sometimes contains traces of selenium or tellurium (Specimens Nos. 60856 and 60864, U.S.N.M.). Origin wild mode of occurrence. Sulphur deposits of such extent as to be of economic importance occur as a product of volcanic activity, or result from the alteration of beds of gypsum. On a smaller scale, and of interest from a purely mineralogical standpoint, are the occur- rences of sulphur through the alteration of pyrite and other metallic sulphides. As a product of volcanic action sulphur is formed through the oxida- tion of hydrogen disulphide (H 2 S), which, together with steam and other vapors, is a common exhalation from volcanic vents and solf ataras. Such deposits on a small scale may be seen incrusting f umaroles in the Roaring Mountain (Specimen No. 72872, U.S.N.M.) or associated with the sinter deposits of the Mammoth Hot Springs in the Yellowstone Park (Specimen No. 72877, U.S.N.M.). It may also be produced through the mutual reaction of hydrogen disulphide (H 8 S) on sulphuric anhydride (SO 3 ), the product being sulphur (S) and water (H 2 O) as THE NONMETALLIC MINERALS. 175 before. To these types belong the sulphur deposits of Utah, Cali- fornia, Nevada, and Alaska in the United States, as well as those of Mexico, Japan, Iceland, and other volcanic regions. Sulphur is derived from the sulphate of lime (gypsum or anhydrite) through the reducing action of organic matter. The sulphate, through the loss of its oxygen, becomes converted into a sulphide, which, through the carbonic acid in the air and water, becomes finally reduced to hydrogen disulphide with the formation of calcium carbonate. According to Fuchs and De Launay 1 there is formed at the same time with the hydrogen disulphide a polysulphide, which in its turn yields a precipitate of sulphur and carbonate of lime. The maximum amount of sulphur which would thus result from the decomposition of a given amount of gypsum is stated to be 24 per cent. This method of origin is illustrated in the celebrated deposit of Sicily, where we have the sulphur partially disseminated through and partly interbedded with a blue-gray limestone. (See Specimen No. 60932, U.S.N.M.). Beneath the sulphur beds as they now exist are found the older gyp- seous beds, which through decomposition have yielded the materials for the lime and sulphur beds now overlying. With these Sicilian sulphurs occur a number of beautiful secondary minerals, as celestite (Specimens Nos.. 60866, 60869, 60877, U.S.N.M.), calcite (Specimens Nos. 60854, 60865, 60871, U.S.N.M.), aragonite (Specimen No. 60859, U.S.N.M.), and selenite (Specimen No. 60857, U.S.N.M.). Sulphur derived directly from metallic sulphides is of little economic interest. Kemp states 2 that masses of pyrite in the calciferous strata on Lake Champlain may yield crusts of sulphur an inch or so thick, and it is not uncommon to find small crystals of the mineral resulting from the alteration of galena, as described by George H. Williams 3 at the Mountain View (Maryland) lead mine. The minute quantities of sulphur found in marine muds are regarded by J. Y. Buchanan* as due to the oxidation of metallic sulphides, which are themselves produced by the action of animal digestive secre- tions on preexisting sulphates, mainly of iron and manganese. Localities. The principal localities of sulphur known in the United States are, in alphabetical order: Alaska, California, Idaho, Louisi- ana, Nevada, Texas, Utah, and Wyoming. With the possible excep- tion of those of Idaho and Texas, and that of Louisiana, these may all be traced to a solfataric origin. The Alaskan deposit, 5 according to Dall, are best developed on the islands of Kadiak and Akutan. 1 Trait6 des Gites Mineraux et Metalliferes, I, p. 259. 2 The Mineral Industry, II, 1893, p. 585. 3 Johns Hopkins University Circulars, X, 1891, p. 74. Proceedings of the Royal Society of Edinburgh, XVIII, 1890-91, p. 17. 5 Alaska and its Resources, Boston, 1870. 176 REPORT OF NATIONAL MUSEUM, California deposits have in times past been worked at Clear Lake, in Modoc County, in Colusa County, in Tehama County (Specimen No. 30118, U.S.N.M.), and in Napa County (Specimen No. 67697,U.S.N.M.). The Louisiana deposits lie in strata of Quaternary age, and are derived from gypsum. The following facts relative to this deposit are from Professor Kemp's paper, already alluded to : Probably the richest and geographically the most accessible of the American localities is in southwestern Louisiana, 230 miles west of New Orleans and 12 miles from Lake Charles. The first hole which revealed this sulphur was sunk in search of petroleum, of which the presence of oil and tarry matter on the surface were regarded, quite justly, as an indication. While more or less of these bituminous substances were revealed by the drill, the great bed of sulphur is the main object of interest. A number of holes have since been put down with the results recorded below, and they leave no doubt that there is a very large body which awaits exploi- tation. The first explorations were made by the Louisiana Petroleum and Coal Oil Company. It was succeeded by the Calcasieu Sulphur and Mining Company. The Louisiana Sulphur Mining Company followed, and now the owners are the American Sulphur Company. The records of four holes are appended. Nos. 1 and 2 were the first sunk, and were about 150 feet apart. Nos. 2, 3, and 4 were put down in 1886. No. 3 is northwest of No. 1. Records of several of the bore holes that have penetrated the sulphur bed. Strata. Original well No. 1. Granet's Wells. Van Slooten's well No. 5. American Sulphu r Company. No. 2. No. 3. No. 4. No. 6. 350 95 125 32 602 No. 7. No. 8. Clay, quicksand, and gravel Soft rock Sulphur bed, 70 to 80 per cent Gypsum and sulphur Depth of hole 110 108 680 344 84 112 12 426 70 119 6 332 138 45 (o) 345 91 110 57 370 72 126 30 598 499 44 52 () 1,231 552 621 525 603 5% o Stopped in sulphur. Analyses from the large bed in holes No. 2 and No. 3 gave the following: Depth. Sulphur. Depth. Sulphur. HoleNo.2. 428 feet Per cent. 62 Hole No. S. 503 feet 441 feet 70 459 feet 80 549 feet 60 466 feet 83 486 feet 90 91 feet 004 Ieet 98 feet feet 540 feet THE NONMETALLIC MINERALS. 177 The difficulties in development lie in the -quicksands and gravel, which are wet and soft, and in the soft rock (hole 1), which yields sul- phurous waters under a head, at the surface, of about 15 feet. The Nevada deposits occupy the craters of extinct hot springs near Humboldt House. These craters are described by Russell 1 as situated on the open desert, above the surface of which they rise to a height of from 20 to 50 feet. Nearly all of the cones are weathered and broken down, and are all extinct, the water now rising to the surface for miles around. The outer surface of the cones is composed of calcareous tufa and siliceous sinter, forming irregular imbricated sheets that slope away at a low angle from the orifice at the top. The interiors of these structures are filled with crystalline gypsum, that in at least two instances is impregnated with sul- phur. One of the cones has been opened by a cut from the side in such a manner as to expose a good section of the material filling the interior, and a few tons of the sul- phur and gypsurn removed. The percentage of sulphur is small, and the economic importance of the deposit, as shown by the excavation already made, will not war- rant the further expenditure of capital. The cone that has been opened is surrounded on all sides by a large deposit of calcareous and siliceous material, thus forming a low dome or crater, with a base many times as great in diameter as the height of the deposit. These cones correspond in all their essential features with the structures that surround hot springs that are still active in various parts of the Great Basin, thus leaving no question as to their origin. They are situated within the basin of Lake Lahontan, and must have been formed and become extinct since the old lake evaporated away. Sulphur is reported as occurring in the chemically formed deposits that surrounded Steamboat Springs, situated midway between Carson and Reno, Nevada. The conditions at these springs must be very simi- lar to those that existed near Humboldt House at the time the cones containing the sulphur were formed. Sulphur is also said to occur in the Sweetwater Mountains, situated on the boundary between Cali- fornia and Nevada, in latitude 38 30'. The extent and geological relations of these deposits are unknown. Another illustration of sulphur deposits of the volcanic type is that furnished by the Rabbit-Hole Sulphur Mines (Specimen No. 16092, U.S.N.M.). These are located in northwestern Nevada, on the eastern border of the Black Rock Desert, and derive their name from the Rab- bit-Hole Springs, a few miles to the southward. The hills bordering the Black Rock Desert on the east are mainly of rhyolite, with a narrow band of volcanic tufa along the immediate edge of the desert. These beds of tufa are stratified and evidently water-lain, and are identical with tufa deposits that occur over an immense area in Oregon and Nevada. At the sulphur mines the tufas contain angular fragments of volcanic rock, and have been cemented by opal and other siliceous infiltrations since their deposition, so that they now form brittle siliceous rocks, with pebbles and fragments of older rocks scattered through the mass. 1 Transactions of the New York Academy of Sciences, I, 1881-1882, p. 172. NAT MUS 99 12 178 KEPORT OF NATIONAL MUSEUM, 1899. In manv places these porous tufas and breccias are richly charged with sulphur^ which fills all the interstices of the rock and sometimes lines large cavities with layers of crystals 5 or 6 feet in thickness. In the Rabbit-Hole District sulphur has been found in paying quantities for a distance of several miles along the border of the desert, but the dis- tribution is irregular and uncertain, and is always superficial so far as can be judged by the present openings. The sulphur has undoubt- edly been derived from a deeply seated source, from which it has been expelled by heat, and escaping upward along the lines of faulting has been deposited in the cooler and higher rocks in which it is now found, though whether the deposition took place by direct sublimation or through the decomposition of hydrogen disulphide can not now be told with certainty. Judging from the siliceous material that cements the tufas, it is evident that the porous rocks in which the sulphur is now found were penetrated by heated waters bearing silica in solution pre- vious to the deposition of the sulphur. The mines occur in a narrow north-and-south belt along a line, of ancient faulting which is one of the great structural features of the region. The association of faults with sulphur-bearing strata of tufa is here essentially the same as at the Cove Creek Mines, yet to be noted. At the Rabbit- Hole Mines, however, no very recent movement of the ancient fault could be deter- mined. This absence of a recent fault-scarp, together with the fact that the mines are now cold and do not give off exhalations of gas or vapor, shows that the solfataric action at this locality has long been extinct, though at the Cove Creek Mines, mentioned below, the depo- sition is still in progress. According to A. F. Du Faur 1 this Cove Creek (Utah) deposit is in Beaver County, near Millard County line. It was first discovered in 1869, but owing to lack of railroad communications remained undevel- oped until 1883. The region is one of comparatively recent volcanic activity. The sulphur occurs impregnating limestone and slate to such a degree that very pure pieces as large as one foot in diameter are obtainable. It also occurs impregnating a decomposed andesite (Speci- men No. 14921, U.S.N.M.). The Cove Creek mines are situated about 2 miles southeast of Cove Creek fort and to the east of the Beaver road in a small basin near the foot of the Sulphur Mountains, surrounded by low hills, with a narrow ravine opening in the west-northwest direction into the plain. The basin is about 6,000 feet above the level of the sea, while the Sulphur Mountains to the east rise about 2,000 feet higher. The hills surrounding the basin consist mainly of andesite, partly also of a very light white trachyte. As far as explored, the sulphur bed extends at least 1,800 feet by 1,000 feet, and the quantity of sulphur contained therein was estimated transactions of the American Institute of Mining Engineers, XVI, 1888, p. 33. THE NONMETALLIC MINEKALS. 179 by Professor vom Rath, at a time when the bed was not as fully exposed as it now is, to be at least 1,300,000 tons. A curved cut has been made through the sulphur bed near the west- ern end, exposing a vertical wall 34 feet high of rich yellow sulphur. The sulphur extends up to the surface over part of the basin, but is mostly covered with sand or rather decomposed andesite. The sur- face of the deposit is wavy, giving the impression of an agitated mass gradually cooled. The sulphur is partly mixed with sand or gypsum. Most of it is yellow color, while some of it is dark gray, and is called "black sulphur." The deposits of pure sulphur partly resemble the so-called "virgin rock," which is formed as a product of distillation in the sulphur-flower chambers, particularly when distillation goes on too rapidly. Some also resemble the delicate crystals formed on the walls of such chambers; others are like the crystals formed in slowly cooled masses of sulphur. Gases escape in many places in the cut and in the prospect holes, together with water holding salts in solution. At some points also a consider- ably elevated temperature is observed. Of the foreign localities of sulphur, the most noted at present are those of Sicily and Japan. The first-named deposits are described as occur- ring in Miocene strata involv- ing, from below up, sandy marls with beds of salt, limey Fig. 1. marls and lignite, gypsum and limestone impregnated with sulphur, black shales, and micaceous sands. Overlying all tliese is a white, marly Pleocene lime- stone, while below the Miocene is the Eocene nummulitic limestone. The sulphur is found in veinlets and sometimes in larger masses, which ramify through the cellular limestone, as shown in fig. 1 and Specimens Nos. 60932, 60862, 60852, U.S.N.M. The yield in sulphur varies from 8 to 25 per cent, rarely running as high as 40 per cent. Below 8 per cent the rock can not be worked. More or less petroleum and bitumen are found in the mines. . Barite and celestite sometimes accompany the sulphur. The mining regions are in the southern central portion of the island Girgenti and Larcara are the chief centers, The mines are distributed over an area 160 to 170 kilometers (about 100 miles) from east to west, and 85 to 90 kilometers (55 miles) from north to south. They occur in groups around centers, partly because the sulphur-bearing stratum SULPHUR. Sicily, Italy. nen No. 60932, U.S.N.M. 180 REPORT OF NATIONAL MUSEUM, 1899. is not continuous, and partly because the sulphur indications are con- cealed by later deposits. The region, moreover, is much faulted. According to Professor Kemp, the common methods of mining are of the crudest description. In most cases the deposits are reached by steep slopes or circular stairways (" scala"), with wide steps, up which boys laboriously bring the crude rock in baskets or sacks. No mine maps are made, and no precautions taken to work beds on a systematic scale. Timbering or any supports for the roof are not generally thought of. A feeling of distrust prevails between the owners of the land and the operators, and between the latter and the miners. These objectionable features arise partly from the irregular nature and uncertainty of the deposits, partly from excessive subdivision of ownership and ill-adapted property laws, and partly from the local prejudices against innovations. Even in one case where an American and an Englishman in partnership secured the right to work a mine, and set about installing suitable hoisting machinery, they were ham- pered by a lawsuit with the owner because of this innovation, and had a long legal contention to establish their undoubted rights. It is a striking fact that in the new developments in Japan, on a remote island and against great natural difficulties, the most modern methods and management prevail, while in Sicily, in the center of the oldest civilization, these are to a great extent of the crudest. The Japanese sulphur deposits -are all of volcanic origin, and the Abosanobori mine (Specimen No. 61941, U.S.N.M.), in Kushiro village, Kawakami-gori, Kushiro Province, Hokkaido, may be taken as fairly typical. The mine is on a conical-shaped mountain of augite andesite which, on its northern side is open, and looks down upon a plain cov- ered with lava and shut in by the walls of the old crater on the other sides. Sulphur is found in different parts of these walls in massive heaps and sulphur fumes still issue nearly everywhere about the mines. The ore as taken from the mines carries from 35 per cent to 90 per cent of sulphur, which is extracted by steam refining works at Hyocha, some 35 miles north of the mine. 1 Other Japanese localities represented in the collection are the Aroya mines, at Onikobe village, Rikuzen Province (Specimen No. 61945, U.S.N.M.), refined sulphur from the Mitsui Production Com- pany at Tokio (Specimen No. 61944, U.S.N.M.), and the active vol- cano of Icvo-San, in Yezo (Specimen No. 72801, U.S.N.M.). In addition to these localities may be mentioned the following, in alphabetical order: Austria, Celebes, Egypt, France, Greece, Hawaii, Iceland, Italy, Mexico (Specimens Nos. 57136 and 57137 from Popo- catepetl), New South Wales, New Zealand, Peru, Russia, Spain, and the West Indies (Specimen No. 33309, U.S.N.M.). 'The Mining Industry of Japan, by Wada Tsunashiro, 1893. THE NONMETALLIC MINERALS. 181 Extraction and 'preparation. Sulphur rarely occurs in nature in any quantity sufficiently pure for commercial purposes. In freeing it from its impurities three methods are employed: (1) Melting, (2) distillation, and (3) solution. In the first the ore is simply dry washed at a low temperature or treated with superheated steam until the sulphur melts and runs off. Specimen No. 60861 shows the rock after being subjected to this treatment. The first process is extremely wasteful; the second much more economical in the end, but demanding a more expensive plant. A process of fusion in a calcium chloride solution has come into use of late years, and bids fair to yield better results than either of the above. In the distillation process the ore is heated in iron retorts until the sulphur distills off and is condensed in chambers prepared for it. Specimen No. 60860 shows the rock after removal of the sulphur by this process. The product is mostly in the form of "flower of sulphur." The method is expensive, but the resultant sulphur very pure. In the third process mentioned the ore is treated with carbon disulphide, which dissolves out the sulphur and from which it is recov- ered by evaporation. This method, while giving good results, is expen- sive and spmewhat dangerous, owing to the explosive nature of the gases formed. 1 Uses. Sulphur is used mainly for making of sulphuric acid though small amounts are utilized in the manufacture of matches for medici- nal purposes, and in the making of gunpowder, fireworks, insecticides, for vulcanizing india rubber, etc. In the manufacture of sulphuric acid the sulphur is burned to sulphurous anhydride (SO 2 ) on a grate and then conducted with a slight excess of air into large lead-lined chambers and mixed with steam and nitrous fumes, where the SO 2 is oxidized to the condition of SO 3 (sulphuric anhydride) and takes up water from the steam forming H 2 SO t (sulphuric acid). Ordinary roll sulphur is quoted in the current price lists at from 1 to 2 cents per pound. (See also under iron pyrites, p. 190.) BIBLIOGRAPHY. R. PUMPELLY. Sulphur in Japan. Geological Researches in China, Mongolia, and Japan. Smithsonian Contri- butions, XV, 1867, p. 11. I. C. RUSSELL. Sulphur Deposits of Utah and Nevada. Transactions of the New York Academy of Science, I, 1882, p. 168. A. FABER DU FAUR. The Sulphur Deposits of Southern Utah. Transactions of the American Institute Mining Engineers, XVI, 1887, p. 33. The Sulphur Mines of Sicily. Engineering and Mining Journal, XL VI, 1888, p. 174. V. LAMANTIA. Sulphur Mines of Sicily. U. S. Consular Report No. 108, 1889, pp. 146-155. 1 The Mineral Industry, II, 1893, p. 600. 182 REPORT OF NATIONAL MUSEUM, 1899. 3. ARSENIC. This substance occurs native in the form of a brittle, tin-white metal, with a specific gravity of 5.6 to 5.7 and a hardness equal to 3.5 of the scale. On exposure it becomes dull black on the immediate surface. It is found, as a rule, in veins in the older crystalline rocks associated with antimony and ores of gold and silver. Some of the more cele- brated localities for the mineral, as given by Dana, are the silver mines of Freiberg (Specimens Nos. 60924 and 67730, U.S.N.M.), Annaberg, Marienberg, and Schneeberg in Saxony; Joachimsthal in Bohemia; Andreasberg in the Harz; Kapnik and Orawitza in Hungary; Kongs- berg in Norway; Zmeov in Siberia; St. Maria aux Mines, Alsace; Mount Coma dei Darden, Italy; Chanarcillo, Chili; San Augustin, Hidalgo, Mexico, and New Zealand. In the United States it has been found at Haverhill, New Hampshire; Greenwood, Maine; near Lead- ville, Colorado; and on Watson Creek, Frozen River in British Columbia. The arsenic of commerce is, however, rarely obtained from the native mineral, but is prepared by the ignition of arsenical pyrites (FeAs 2 ) or arsenical iron pyrites (FeS 2 ,FeAs 2 ). The white arsenic of commerce (arsenious acid, As 2 O 3 ), though occurring sometimes native as arsenolite in the form of botryoidal and stalactitic crusts of a white or yellowish color, is, as a rule, obtained as a by-product in the metal- lurgical operations of extracting certain metals, particularly cobalt and nickel, from their ores. Such ores as niccolite, a nickel arsenide (NiAs), gersdorffite (NiAsS), Rammelsbergite (NiAs 2 ), Smaltite (CoAs 2 ), Skutterudite (CoAs 3 ), Proustite (Ag 3 AsS 3 ), and other arsen- ides and sulpharsenides on roasting give up their arsenic in the form of fumes, which are condensed in chambers prepared for this purpose. Uses. Arsenic is utilized in the form of arsenious acid (As z O 3 ) in dyeing, calico printing, in the manufacture of various pigments, in arsenical soaps, in the preparation of other salts of arsenic, and as a preservative in museums, particularly for the skins of animals and birds. 4. ALLEMONTITE. Allemontite, or arsenical antimony of the formula SbAs 3 , = arsenic 65.2; antimony 34.8, occurs somewhat sparsely at Allemont in France, Pribram, Bohemia, and other European localities associated with sphalerite, antimony, etc. (Specimen No. 67728, U.S.N.M.). So far as the writer has information the mineral has not as yet been found in sufficient quantity to be of economic value. THE NONMETALLIC MINERALS. 183 II. SULPHIDES AND ARSENIDES. 1. REALGAR. This is a monosulphide of arsenic, AsS, = sulphur 29.9 per cent; arsenic, 70.1 per cent; hardness, 1.5 to 2; specific gravity, 3.55; color, aurora red or orange yellow, streak the same. 2. ORPIMENT; AURIPIGMENT. A trisulphide of arsenic, of the formula As 2 S 3 , = sulphur 39 per cent, arsenic, 61; hardness, 1.5; specific gravity, 3.4 to 3.5. Color, lemon yellow. This mineral occurs usually associated with realgar at the localities mentioned below. Occurrences. Realgar and orpiment are very beautiful, though not abundant minerals which occur associated with ores of silver and lead in various European mining regions and also those of Japan (Specimen No. 11864, U.S.N.M.), Hungary (Specimen No. 66813, U.S.X.M.), Bohemia, Transylvania, and Saxony. They have been reported in the United States in beds of sandy clay beneath lava in Iron County, Utah, and form the so-called "Arsenical gold ore" of the Golden Gate Mine, Mercur, Tooele County, this same State (Specimen Xo. 53363, U.S.N.M.); also in San Bernardino County, California; Douglas County, Oregon (Specimen No. 62101, U.S.N.M.), and in minute quantities in the geyser waters of the Yellowstone National Park. The realgar and orpiment of the Coyote mining district. Iron County, Utah, occur in a compact, sandy clay, occupying a horizontal seam or layer about 2 inches thick, not distinctly separated from the clay, but lying in its midst in lenticular and nodular masses. The bulk of the layer consists of realgar in divergent, bladed crystals, closely and confusedly aggregated, sometimes forming groups of brilliant crystal- line facets in small cavities toward the center of the mass. The orpi- ment is closely associated with the realgar in the form of small and delicately fibrous crystalline rosettes, and small spherical aggregations made up of fine radial crystals, and also in bright yellow, amorphous crusts in and around the mass of the realgar. Fine parallel seams of gypsum occur both above and below the layer, and the strata of arena- ceous clays above for 30 feet or more are charged with soluble salts which exude and effloresce upon the surface of the bank, forming hard crusts. The whole appearance and association of the minerals indi- cates that they have been formed by aqueous infiltration since the deposition of the beds. 1 Orpiment is said 2 to occur at Tajowa, near Xeusohl, Hungary, as nodular masses and isolated crystals in clay or calcareous marl. 1 W. P. Blake, American Journal of Science, XXI, 1881, p. 219. 2 H. A. Miers, Mineralogical Magazine, July, 1892, p. 24. 184 KEPORT OF NATIONAL MUSEUM, 1899. . Realgar is used mainly in pyrotechny, yielding a very bril- liant white light when mixed with saltpeter and ignited. It is now artificially prepared by fusing together sulphur and arsenious acid. 1 Orpiment is used in dyeing and in preparation of a paste for removing hair from skins. According to the British consular reports there were exported from Baghdan, in 1897, some 55,600 pounds of the mineral for use as a pigment. As with realgar, the mineral is now largely prepared artificially. The name "orpiment" is stated by Dana to be a corruption of auripigment, golden paint, in allusion to the color. BIBLIOGRAPHY. W. I*. BLAKE. Occurrence of Realgar and Orpiment in Utah Territory. American Jourual of Science, XXI, 1881, p. 219. H. B. FULTON. Arsenic in Spanish Pyrites, and its elimination in the local treat- ment for production of copper precipitate. Journal of the Society of Chemical Industry, V, 1886, p. 296. Production of Arsenic in Cornwall and Devon. Engineering and Mining Journal, LII, 1891, p. 96. WILLIAM THOMAS. Arsenic. The Mineral Industry, II, 1893, p. 25. 3. COBALT MINERALS. Several minerals contain cobalt as one of their essential constituents in sufficient quantity to make them of value as ores. In other cases the cobalt exists in too small quantities to pay for working for this substance alone, and it is obtained as a by-product during the process of extraction of other metals, notably of nickel. The common cobalt- bearing minerals, together with their chemical composition, mode of occurrence, and other characteristics are given below : COBALTITE. Cobaltine, or cobalt glance. (Specimens Nos. 60922, 34266, U.S.N.M.) This is a sulpharsenide of cobalt of the formula Co AsS, = Sulphur 19.3 per cent; arsenic, 45.2 per cent; cobalt, 35.5 per cent; hardness 5.5, and specific gravity 6 to 6.3. The luster is metallic and color silver white to reddish. When in crystals, commonly in cubes or pyritohedrons. Analysis of a massive variety from I, Siegen, Westphalia; II, Skutterud, Norway, and III and IV, Daschkessan, in the government of Elizavetpol, Caucasus, as given by various author- ities, yielded results as below: Constituents. I. ii. in. IV. Arsenic. Sulphur 19.35 20 08 Cobalt 33 71 Iron Nickel 22 26 Undetermined... 44 26 1 Wagner's Chemical Technology, p. 87. THE NONMETALLIC MINERALS. 185 In Saxony the mineral (Specimens Nos. 60922 and 67736, U.S.N.M.) occurs in lodes in gneiss and in which heavy spar (baryte) forms the characteristic gangue. It is associated with other metallic sulphides, notably those of lead and copper. At Skutterud and Snarum, Nor- way, the cobaltiferous fahlbands, according to Phillips 1 Occur in crystalline rocks varying in character between gneiss and mica schists, but from the presence of hornblende they sometimes pass into hornblende schists; among the accessory minerals are garnet, tourmaline, and graphite. These schists, of which the strike is north and south, and which have an almost perpendicular dip, contain fahl- bands very similar in character to those of Kongsberg. They differ from those of that locality, however, inasmuch as while here the fahlbands are often sufficiently impreg- nated with ore to pay for working, those of Kongsberg, although to some extent contain- ing disseminated sulphides, are only of importance as zones of enrichment for ores occurring in veins. The ore zones usually follow the strike and dip of the surround- ing rocks, and vary in breadth from 2 to 6 fathoms. The distribution of the ores is by no means equal, since richer and poorer layers have received special names and are easily recognized. The Erzbander, or ore bands, are distinguished from the Reicherzbander, or rich ore bands, while the bands of unproductive rock are known as Felsbander. The predominant rock of the fahlbands is a quartzose granular mica schist, which gradually passes into quartzite, ordinary mica schist, or gneiss. The ores worked are cobalt glance, arsenical, and ordinary pyrites containing cobalt, skutterudite, magnetic iron pyrites, copper pyrites, molybdenite, and galena. It is remarkable that in these mines nickel ores do not accompany the ores of cobalt in any appreciable quantity. The principal fahlband is known to extend for a distance of about 6 miles, and is bounded on the east by a mass of diorite which protrudes into the fahlband, while extending from the diorite are small dikes or branches traversing it in a zigzag course. It is also intersected by dikes of coarse-grained granite which contain no ore, but which penetrate the diorite. The Skutterud mine in 1879 produced 7,700 tons of cobalt ore, which yielded 108 tons of cobalt schlich (concentrates), containing from 10 to 11 per cent of cobalt, and worth about 11,000. At Dacshkessan the ore occurs under a sheet of diabase, the cobaltite being in the wall rock of this sheet, and which carries also garnets and copper pyrites. In 1887, 1,216 kilograms of the mineral were extracted; in 1888, 928 kilograms, and in 1889, 12,960 kilograms, besides some 3,000 kilograms of cobaltiferous matter obtained in treating the cobal- tiferous copper ores. 2 SMALTITE. (Specimen No. 66757, U.S.N.M.) This is essentially a cobalt diarsenide of the formula CoAs 2 , = arsenic, 71.8 per cent; cobalt, 28.2 per cent; hardness, 5.5 to 6; specific gravity, 6.4 to 6.6. Color, white to steel gray. Through the assumption of nickel the min- eral passes by gradations into chloanthite. J Ore Deposits, by J. A. Phillips, p. 389. 2 Annales des Mines, II, 1892, p. 503. 18 g KEPOET OF NATIONAL MUSEUM, 1899. Analyses of samples from (I) Schneeberg, Saxony, and (II) Gunnison County, Colorado, as given by Dana, yielded results as below: Constituents. II. Areenic 1 71.53 63.82 ArS>eU1L n OQ Sulphur ; 1 '* i Cobalt , 18-07 1.55 11.59 Iron - 7 ' 31 v . Vpl ; 1.02 15.99 Trace. Copper 0.16 The mineral occurs like cobaltite in veins associated with other metallic arsenides and sulphides. SKUTTERUDITE is the name given to a cobaltic arsenide of the formula CoAs 3 , = arsenic^ 79.3; cobalt. 20.7. It is of a tin- white color, varying to lead-gray, has a hardness of 6, and specific gravity of 6.72 to 6.86. It occurs associated with cobaltite, titanite, and hornblende in a vein in gneiss at Skutterud, Norway, The name safflorite is given to a cobalt diarsenide closely resembling smaltite but differing in being orthorhombic, rather than isometric in crystallization. The composi- tion as given by Dana is quite variable, running from 61 per cent to 70 per cent arsenic, and 10 to 23 per cent cobalt, with 4 to 18 per cent of iron and smaller amounts of sulphur, copper, nickel, and bismuth. It is found associated with smaltite in various localities. GLAUCODOT is a sulpharsenide of cobalt and iron of the formula (Co, Fe) AsS, = sulphur, 19.4 per cent; arsenic, 45.5 per cent; cobalt, 23.8 per cent; iron, 11.3 per cent. Color, grayish; hardness, 5; specific gravity, 5.9 to 6. Actual analysis of a Chilean variety yielded (accord- ing to Dana) As 43.2, S 20.21, Co 24.77, Fe 11.90. It is therefore essen- tially a ferriferous cobaltite, that is, a cobaltite in which a part of the cobalt has been replaced by iron. The mineral is found at Huasco, Chile, associated with cobaltite in a chloritic schist. The name allo- clasite is given to a variety of glaucodot containing bismuth and answering to the formula Co (As, Bi) S. The composition as given is somewhat variable. Arsenic, 28 to 33 per cent; bismuth, 23 to 32 per cent; sulphur, 16 to 18 per cent; cobalt, 20 to 24 per cent; iron, 2.7 to 3.8 per cent. It is reported only from Orawitza, Hungary. LINN^EITE (Specimens Nos. 56159, 65309, U.S.N.M.) is a sulphide of cobalt with the formula Co 3 S 4 , = sulphur, 42.1per cent; cobalt, 57.9 per cent; a part of its cobalt is commonly replaced by nickel, giving rise to its variety siegenite. The mineral is brittle, of a pale steel- gray color, tarnishing red. Hardness, 5.5 and specific gravity 4.8 to 5. When crystallized it is commonly in octahedrons. The fol- lowing analyses of a nickel-bearing variety (siegenite) are quoted from Dana: THE NONMETALLIC MINEKALS. 187 Constituents. S. CO. Ni. Fe. Cu. Miisen, Prussia 41 00 43 86 5 31 4 10 Mineral Hill, Maryland Mine La Motte Missouri 39.70 41 54 25.68 21 34 29.56 30 53 1.% 3 37 2.23 The mineral occurs in gneiss in Sweden; with barite and siderite at Miisen; in limestone with galena and dolomite at Mine La Motte, Missouri, and with sulphides of iron and copper in chloritic schists in Maryland. SYCHNODYMITE has the formula (Co, Cu) t S 5 , and yields sulphur, 40.64 per cent; copper, 18.98 per cent; cobalt, 35.79 per cent; nickel, 3.66 per cent; iron, 0.93 per cent. It is of a steel-gray color, metallic luster, and has a specific gravity of 4.75. ERYTHRITE or COBALT BLOOM (Specimens Nos. 17698, 51909, 56463, 53096, and 67759, U.S.N.M.) is the name given to a hydrous cobalt arsenate of the formula Co 3 As 2 O 8 -h8H 2 O, = arsenic pentoxide, 38.4 per cent; cobalt protoxide, 37.5 per cent, and water, 24.1 per cent. It occurs in globular and reniform shapes and earthy masses of a crimson to peach-red color associated with the arsenides and sulphar- senides mentioned above and from which it is derived by a process of oxidation. In Churchill County, Nevada, it occurs as a decomposition product of a cobalt bearing niccolite. It is also found at the Kelsey mine, Compton, in Los Angeles County, California; associated vith cobaltite at Tambillo and at Huasco, Chile, and under similar con- ditions in various p..rts of Europe. ASBOLITE, or earthy cobalt (Specimen No. 60993, U.S. N.LI.), is a black and earthy ore of manganese (wad) which sometimes carries as high as 30 per cent of cobaltic oxide. It takes its name from the Greek ctfffioXaivG), to soil like soot. ROSELITE is an arsenate of lime, mag- nesia and cobalt with the formula (Ca, Co, Mg) 3 As 2 O g , 2H 2 O, = arsenic pentoxide, 51.4 per cent; lime, 28.1 per cent; cobalt protoxide, 12.5 per cent; water, 8 per cent. It is of a light to dark rose-red color, hardness 3. 5 ; specific gravity 3. 5 to 3. 6, and vitreous luster. SPH^ROCOBALTITE is a cobalt protocarbonate of the formula CoCO 3 , = carbon dioxide, 37.1 per cent; cobalt protoxide, 62.9 per cent. It is also of a rose-red color, varying to velvet black. Hardness 4, and specific gravity 4.02 to 4.13. It occurs but sparing, associated with roselite at Schneeberg in Saxony. REMINGTONITE is a hydrous carbonate the exact composition of which has not been ascertained. COBALTOMENITE is a supposed selenide of cobalt. BIEBERITE, or cobalt vitriol, is a sulphate of the formula CoSO 4 + 7 H 2 O. The color is flesh to rose red. It is soluble in water, has an astringent taste, and occurs in secondary stalactitic form. Pateraite is a possible molybdate of cobalt. Aside from the possible sources mentioned above, cobalt occurs 188 REPORT OF NATIONAL MUSEUM, 1899. very constantly associated with the ores of nickel (niccolite, millerite, chloanthite, etc.), and is obtained as a by-product in smelting. Con- siderable quantities have thus from time to time been obtained from the Gap mines of Pennsylvania, Mine La Motte, Missouri, and Love- lock, Nevada. (Specimen No. 61324, U.S.N.M.) The nickel mines of New Caledonia are perhaps the most productive. The ore here (a sili- cate), carries some 3 per cent of cobalt protoxide. (Specimen No. 61027, U.S.N.M.) A vein of cobalt ore near Gothic, Gunnison County, Colorado, is described as lying in granite, the gangue material being mainly cal- cite, throughout which was disseminated the ore in the form of smaltite. With it were associated erythrite, a small amount of iron pyrites, and native silver. An analysis of this ore yielded as below: Bismuth 1. 13 Copper 0. 16 Nickel Trace. Silver... .. Trace. Cobalt 11.59 Iron 11.99 Arsenic; 63. 82 Silica 2. 60 Lead 2. 05 Sulphur 1. 55 A cobalt ore, consisting of a mixture of glaucodot and erythrite, occurring near Carcoar Railway Station, New South Wales, has the composition given below: 94. 89 Constituents. ' II. Moisture Metallic arsenic .120 51. 810 2.180 29.010 Metallic cobalt .' Metallic nickel 10.447 .590 13.830 390 Metallic iron Alumina ... . Metallic manganese \il Nil Metallic calcium Nil 71 Magnesium 1.480 .22 Gold Silver Sulphur Gangue (insoluble in acids)... . 22 078 26 31 Specific gravity 99.905 5 43 99.67 According to the Annual Report, Department of Mines, for 1888, this ore occurs concentrated in irregular hollows and bunches, often intimately mixed with diorite in a line of fissure between an intrusive diorite and slate, the fissure running for some distance following the line of junction between the two rocks, and being presumably formed at the time of the extrusion of the diorite. THE NONMETALLIC MINEEALS. 189 Other cobalt ores, carrying from 13 to 15 per cent of cobalt oxide, occur near Nina. 1 Uses. Cobalt is produced and sold in the form of oxide and used mainly as a coloring constituent in glass and earthen wares. Only some 200 tons are produced annually the world over. The market value of the material is variable, but averages about $2 a pound. BIBLIOGRAPHY. Fuchs et De Launay, Traite des Gites Mineraux, II, pp. 75-91. 4. ARSENOPYRITE; MISPICKEL; OR ARSENICAL PYRITES. Composition. Somewhat variable. Essentially a sulpharsenide of iron of the formula FeAsS, or FeS 2 , FeAs 2 ,= arsenic, 46 per cent; sulphur, 19.7 per cent, and iron, 34.3 per cent. The name danaite is given to a cobaltiferous variety. The specific gravity of the min- eral varies from 5.9 to 6.2. Hardness, 5.5 to 6. Colors, silver white to steel gray, streak dark gray to black; luster, metallic. Brittle, Occurrence. The mineral occurs principally in crystalline rocks, and is a common associate of ores of silver, gold, tin, and lead. It is at times highly auriferous, forming a valuable ore of gold, as -in New South Wales and more rarely in California and Alaska. It is found in nearly all the States bordering along the Appalachian Mountain system, but in no instance is regularly mined excepting incidentally in the process of working other metals. Concerning its occurrence abroad Dana states that it is "abundant at Freiberg and Munzig, where it occurs in veins (Specimens Nos. 62803, 66809, 66810, 73104, U.S.N.M.); at Reichenstein in Silesia in serpentine; at Auerbach in Baden; in beds at Breitenbrunn and Raschau, Andreasberg and Joachimsthal; at Tunaberg in Sweden; at Skutterud in Norway; at Wheal Mawdlin and Unanimity, Cornwall, and at the Tamar mines in Devonshire, England (Specimens Nos. 67456, 67457, U.S.N.M.) and in Bolivia. Uses. The only use of the mineral is as an ore of arsenic. 5. LOLLINGITE; LEUCOPYRITE. The prismatic arsenical pyrites, or leucopyrite, is essentially a diar- senide of iron, with the formula FeAs 2 , though usually contaminated with a little sulphur and not infrequently cobalt, bismuth, or antimony. It has a specific gravity of 7 to 7.4, hardness of 5 to 5.5, metallic luster and silver-white to steel-gray color. The mineral has been found at Edenville, New York (Specimen No. 67744, U.S.N.M.); Roxbury, Connecticut, and other places in the United States and associated with other arsenides and sulpharsenides in the gold and silver mines of Europe. 1 Complete analyses of these are given in Catalogue of the New South Wales Exhibit, World's Columbian Exposition, Chicago, 1893, p. 330. 190 REPORT OF NATIONAL MUSEUM, 1899. 6. PYRITES. Two forms of the disulphide of iron are common in nature. The first, known simply as pyrite or iron pyrites, occurs in sharply denned cubes and their crystallographic modifications (Specimen No. 51740, U.S.N.M.). or in granular masses of a brassy -yellow color (Specimen No. 62152, 'u.S.N.M.). The second, identical in composition, crystallizes in the othorhombic system (Specimens Nos. 17124, 55206, and 73613,U.S.N.M.), but is more common in concretionary (Specimen No. 62976, U.S.N.M.), botryoidal (Specimen No. 30772, U.S.N.M.), and stalactitic (Specimens Nos. 62800 and 67761, U.S.N.M.) forms, which are of a dull grayish-yellow color. This form is known as the gray iron pyrites. Both forms have the chemical composition, FeS 2? =iron 46.6 per cent and sulphur 53.4 per cent. The ore as mined is. however, never chemically pure, but contains admixtures of other metallic sulphides, besides, at times, considerable quantities of the precious metals. The following analyses l of materials from well-known sources will serve to show the general variation: Constituents. I. II. III. IV. V. VI. VII. } Sulphur 48.0 48.0 48.02 40.00 47.76 46.40 45.60 Iron .. 43.0 44.0 42.01 35.0 43.99 29.00 88.92 1.6 16 4.00 3.69 1.5t Zinc Silica 1.5 1.5 5.0 3.7 7.60 20.00 0.24 1.99 9.25 3.75 6.00 8.70 Trare ... Trace. 0.83 0.10 Trace. Silver and gold... Lead , 1 i Trace. Trace. ' 010 64 I. Milan, Coos County, New Hampshire; II. Ro we, Massachusetts; III. Louisa County. Virginia; IV. Sherbrooke, Canada; V. Rio Tinto, Spain; VI. near Lyons, France: VII. Westphalia, Germany. Pyrite is sufficiently hard to scratch glass, and this, together with its color, crystalline form, and irregular fracture, is sufficient for its ready determination in most cases. Once known, it is thereafter readily rec- ognized. Owing to its yellow color, the mineral has by ignorant per- sons been mistaken not infrequently for gold which, however, it does not at all resemble and has hence earned the not very flattering but quite appropriate name of fool's gold." In certain cases, however, it carries the precious metals, and in many regions is sufficiently rich in gold to form a valuable ore. Jfode of occurrence. Pyrite is one of the most widely disseminated of minerals, both geologically and geographically, occurring in rocks of all kinds and of all ages the world over. It is found in the form of 1 Mineral Resources of the United States, 1883-1884, p. 877. THE NONMETALLIC MINERALS. 191 disseminated grains throughout the mass of a rock, or along the line of contact between basic eruptivesand sedimentaries; as irregular and sporadic and concretionary masses in sedimentary rocks and modern sands and gravels; in the form of true fissure veins, and as interbedded, often lenticular masses, sometimes of immense size, lying conformably with the stratification (or foliation) of the inclosing rock. On the immediate surface the mineral is in most cases considerably altered by oxidation and hydration, forming the caps of gossan or limonite. The origin of the mineral in the older crystalline rocks, as that of the rocks themselves, is not infrequently somewhat obscure. In sedi- mentary rocks it is undoubtedly due to the precipitation of the included ferruginous matter by sulphureted and deoxidizing solutions from decomposing animal and vegetable matter. Some of the pyritiferous deposits, as those of Louisa County, Virginia (Specimens Nos. 54239, 54241, and 54242, U.S.N.M.), and Huelva. Spain, are of enormous proportions. The first named is described 1 as over 2 miles in length, and to have been exploited to upwards of 600 feet in depth and in width, from foot to hanging rock, as high as 60 feet of pure ore (see large Specimen No. 54242, U.S.N.M.). The average width of the two worked beds is upward of 18 feet. The rocks inclosing the deposits consist principally of talcose and hydromica slates. At Rio Tinto the ore is described 2 as occurring in immense masses several thousand feet in length and from 300 to 800 feet in width, extending in depth to an unknown distance. The ore (Specimen No. 11427, U.S.N.M.) is very clean and massive, containing besides sulphur and iron only some 2 to 4 per cent of copper and traces of silver and gold. The material is mined wholly from open cuts and to a depth of some 400 feet. The country rock is described as of Silurian and Devonian schists near contact with diorites. Uses. With the exception of the small amount utilized in the prep- aration of vermilion paints and the still smaller amount used for jewelry, almost the sole value of the mineral is for the manufacture of sulphuric acid and the sulphate of iron, known as green vitriol or cop- peras. In the process of making sulphuric acid the ore is roasted or burnt in specially designed ovens and furnaces until the mineral is decomposed, the sulphur fumes being caught and condensed in cham- bers prepared for the purpose. By the Glover and Ga^v-Lussac method from 280 to 290 parts of sulphuric acid of a density of 66 Baume may be obtained for each 100 parts of sulphur in the ore or about 2,565 pounds of acid to 1 ton (2,000 pounds) of average ore. In the manufacture of copperas the ore is broken into small pieces and thrown into piles over which water is allowed to drip slowly. A 1 Origin of the Iron Pyrites Deposits in Louisa County, Virginia, by F. L. Nason, Engineering and Mining Journal, LVII, 1894, p. 414. - A Visit to the Pyrite Mines of Spain, Eng. and Min. Jour., LVI, 1893, p. 498. 192 REPORT OF NATIONAL MUSEUM, 1899. natural oxidation takes place, whereby the sulphide is transformed into a hydrated sulphate. The latter being soluble runs off in solution in the water, which must be collected and evaporated in order to obtain the salt. Thus prepared the sulphate is used in dyeing, in the manu- facture of writing ink, as a preservative for wood, and as a disinfectant. It has also been used in the manufacture of certain brands of fertilizers. The method of manufacture as formerly carried on at Strafford, Vermont, is given below: The process consists in first raising the ore from the bed, which is principally done with the help of gunpowder. The blocks of ore are then broken up into small pieces, to facilitate the decomposition, by suffering the oxygen contained in water and the atmosphere to come more directly in contact with the material composing the ore. Large heaps of these pieces, called leaches, are made upon a tight plank bottom or upon a sloping ledge of solid rock, where the liquor or lye that subsequently runs from them may be saved. In dry weather a small stream of water is made to flow upon and penetrate these leaches in order to produce a spontaneous combustion, which in warm weather com- mences in a few days, and if properly managed will continue several weeks. When combustion is taking place great care is requisite in order to have the work go on suc- cessfully, for if too much water is suffered to penetrate the leach or heap the decom- position is checked by the reduction of temperature and the lye or liquor issuing from it is too weak to be valuable, and if there is not water enough put on the leach the decomposition is also arrested by the absence of the oxygen found in the water, which is necessary to convert the sulphurous acid into the sulphuric, that sulphate of iron or copperas may be produced. The liquor that runs from the leaches is collected in reservoirs, from which it can be taken at pleasure. Below the reservoirs upon the hillside buildings are erected, called evaporators, to which liquor is conducted in troughs from the reservoirs in small streams that are divided and subdivided by means of perforated troughs, brush, etc. Several tiers of brush are arranged in the building, through which the liquor is made to pass to facilitate the process of evaporation. In dry, windy weather the evaporation is oftentimes so rapid that the brush and other substances with which the liquor comes in contact during the latter part of its journey often have an incrustation of copperas formed upon them; but upon the return of rainy weather the humid atmosphere checks the evaporation, and the crust of copperas is dissolved and passes with the liquor into reservoirs prepared to receive it. The liquor, which is now very strongly impregnated with copperas, is conducted into leaden boilers, where heat is applied and the liquor redi-ced to a strength indi- cated by the acidimeter to be right for the production of copperas. The liquor is then placed in vats of lead or of brick and water cement, called crystallizers, and after remaining from eight to ten days a crust of copperas is formed upon the bottom and sides of the vats, composed of nicely formed crystals. The water remaining in the crystallizers is then pumped back into the boilers, the crust of copperas removed, and, after being sufficiently drained, it is packed in casks ready for market. 1 [See also under Alum shale and vitriol stone, p. 421.] The analyses given below show (1) the composition of fresh pyrite from the Coal Measures of Mercer County, Pennsylvania, and (2) and (3) that of two varieties of paint produced from it by calcination. 2 Geology of Vermont, II, 1861, p. 830. 2 Report M. M. Second Report of Progress in the Laboratory of the Survey at Har- risburg, Second Geological Survey of Pennsylvania, 1879, p. 374. V THE NONMETALLIC MINEBALS. 193 Constituents. 1. 2. 3. Bisulphide of iron Bisulphide of copper 96.161 Trace. 0.415 0.405 Sesquioxide of iron. 66 143 77 143 Alumina .653 .697 .543 6 800 5 142 Lime .450 .160 160 140 100 100 Silica .680 3.880 3.980 Sulphuric acid 13. 110 7.334 Water and carbonaceous matter Undetermined 1.916 9.195 5.194 Total 100 000 100 000 100 000 Pyrite on decomposing in the presence of moisture in the ground sometimes gives rise to an acid sulphate of iron. This may attack aluminous minerals when such are present, giving rise thus to solutions of sulphate of iron and alumina, which come to the surface as "alum springs," or, if no alumina is present, merely as iron or chalybeate springs, which are of more or less medicinal value. The presence of such sulphates in a soil is readily detected by the well-known astrin- gent taste of green vitriol and alum, even where the quantity is not sufficient to appear as a distinct efflorescence. Impregnation of these salts in soils are by ignorant persons sometimes assumed to be of great medicinal value, and the writer has in mind a case in one of the Southern States, in which the aqueous leachings of such a soil were regularly bottled and sold as a specific for nearly all the ills to which the flesh is heir, though prescribed especially for flux, wounds, and ulcers. (See also under Alum, p. 416.) BIBLIOGRAPHY. W. H. ADAMS. The Pyrites Deposits of Louisa County, Virginia. Transactions of the American Institute of Mining Engineers, XII, 1883, p. 527. WILLIAM MARTYN. Pyrites. Mineral Resources of the United States, 1883-84, p. 877. J. H. COLLINS. The Great Spanish Pyrites Deposits. Engineering and Mining Journal, XL, 1885, p. 79. E. 1). PETERS. A Visit to the Pyrites Mines of Spain. Engineering and Mining Journal, LVI, 1893, p. 498. FRANK L. NASON. Origin of the Iron Pyrites Deposits in Louisa County, Virginia. Engineering and Mining Journal, LVII, 1894, p. 414. M. DRILLON. The Pyrites Mines of Sain-Bel. Minutes of Proceedings of the Institute of Civil Engineers, CXIX, 1894-95, p. 470. 7. MOLYBDENITE. A disulphide of molybdenum having the formula MoS 2 , = sulphur 40 per cent, molybdenum 60 per cent. NAT MUS 99 13 194 KEPOBT OF NATIONAL MUSEUM, 1899. This mineral, like graphite, occurs, as a rule, in small, black, snmmg scales, sometimes hexagonal in outline and with a bright metallic luster. It is soft enough to be readily impressed with the thumb nail, and leaves a bluish-gray trace on paper. On porcelain it leaves a lead gray, slightly greenish streak. This faint greenish tinge, together with its property of giving a sulphur reaction when fused with soda, furnish a ready means of distinguishing it from graphite, which it so closely resembles. Through alteration it sometimes passes over into molybdite or molybdic ocher, a straw-yellow to white ocherous mineral of the formula MoO 3 , = oxygen 33.3 per cent, molybdenum 66.7 per cent. Occurrence. The mineral has a wide distribution, occurring in embedded masses and disseminated scales in granite (Specimen No. 62169, U.S.N.M.), gneiss, syenite, crystalline schists, quartz (Specimen No. 60995, U.S.N.M.), and 'granular limestone. It is found in Nor- way, Sweden, Russia, Saxony, Bohemia, Austria, France, Peru, Brazil, England, and Scotland, throughout the Appalachian region in the United States and Canada (Specimen No. 53046, U.S.N.M.), and in various parts of the Rocky and Sierra Nevada mountains. In Okan- ogan County, Washington, the mineral occurs in beautiful large flakes in an auriferous quartz vein traversing slates. (Specimen No. 53126, U.S.N.M.) On Quetachoo-Manicouagan Bay, on the north side of the Gulf of St. Lawrence, the mineral is reported 1 as occurring disseminated in a bed of quartz 6 inches thick, in the form of nodules from 1 to 3 inches in diameter and in flakes which are sometimes 12 inches broad by i inch in thickness. Molybdenite is also found in the form of finely disseminated scales or small bunches among the iron ores of the Hude mine at Stanhope, New Jersey, sometimes constituting as high as 2 per cent of the ore. Molybdenum is also a constituent of the mineral wulfenite, or molybdate of lead. Uses. The principal use to which molybdenite has as yet been put is in the preparation of molybdates for the chemical laboratory. It is stated that a fine blue pigment can be prepared from it, which it has been proposed to use as a substitute for indigo in dyeing silk, cotton, and linen. The metal molybdenum is produced but rarely and only as a curiosity, and has a purely fictitious value. Up to the present time there has been no constant demand for the mineral nor regular source of supply. 1 Geology of Canada, 1863, p. 754. THE NONMETALLIC MINERALS. 195 HI. HALIDES. 1. HALITE; SODIUM CHLOKIDE; OR COMMON SALT. Composition Na Cl,= sodium 60.6 per cent; chlorine 39.4 per cent. The natural substance is nearly always more or less impure, as noted later. Hardness, 2.5; specific gravity, 2.1 to 2.6 per cent. Colorless or white when pure, but often yellowish or red or purplish by the presence of metallic oxides and organic matter. Readily soluble in cold water, and has a saline taste. Crystallizes in the isometric system, Fig. 2. CLUSTER OP HALITE CRYSTALS. Stassfurt, Germany. Specimen No. 40222, U.8.N.M. usually in cubes (fig. 2, Specimen No. 40222, U.S.N.M.), but some- times in octahedrons, the faces of the crystals (particularly when pre- pared artificially) being often cavernous or hopper shaped. Sometimes occurs in fibrous forms, which it has been suggested are pseudomor- phous after fibrous gypsum (Specimen No. 64733, U.S.N.M.). Often found in the form of massive, crystalline granular aggregates com- monly known as rock salt (Specimens Nos. 67558, 64736, 62946, U.S.N.M.). Sylvite, the chloride of potassium, sometimes occurs associated with halite, where.it has formed under similar conditions. From halite ig6 REPOBT OF NATIONAL MUSEUM, 1899. it can be distinguished by its crystalline form, that of a combination of cube and octahedron (Specimen No. 40223, U.S. KM. See fig. 4, p. 203), and more biting taste. Owing to its ready solubility it is rarely found in a state of nature. Bischofite, the chloride of mag- nesium (Specimen No. 62428, U.S.N.M.) is still more soluble and practically unknown except in crystals artificially produced. Origin and occurrences. Sodium in the form of chloride, to which is commonly given the simple name of salt, is one of the most widely disseminated of natural substances, and not infrequently occurs in large masses interstratified with other rocks of the earth's crust in such a manner as to constitute a true rock mass. The geological history of these beds of rock salt is as follows: No terrestrial waters are absolutely pure, but all hold in solution more or less mineral matter which has been taken up from the rocks and soils with which they have come in contact. The nature of these impurities depends on the nature of the formations permeated and their relative solubility.' Numerous analyses of river waters have shown that the substances mentioned below, though sometimes exist- ing as mere traces, are almost invariably present; these are sodium, potassium, magnesium, silicon, aluminum, and iron, which exist mostly in the form of carbonates, oxides, sulphates, and chlorides. When a stream bearing these substances in solution flows into a lake with no outlet, as the Great Salt Lake or the Dead Sea, the water is returned to the atmosphere by evaporation, while the impurities remain. In this way the water gradually becomes charged more and more heavily with mineral matter, until the point of saturation is reached and further concentration is impossible without precipitation. When such precipitation of mineral, matters takes place, it is in the inverse order of their solubilities; that is, those substances which are least soluble will, under like conditions of temperature, be first precipi- tated. Hence a water containing the ingredients before mentioned on being subjected to complete evaporation would deposit its load in the following order: (1) Carbonates of lime and magnesia in the form of limestones, marls, and dolomites; (2) sulphate of lime in the form of anhydrite and gypsum; (3) chloride of sodium, or common salt; and these followed in regular order by the sulphates of magnesia and soda (Epsom salt and Glauber's salt) and the chlorides of potassium and magnesium. These last are, however, so readily deliquescent that they are rarely found crystallized out in a state of nature as above noted. It rarely happens, however, that nature's processes are sufliciently regular and uninterrupted to allow a complete precipitation of the pure salts as above outlined. During periods of flood suspended silt may be poured into the inclosed basin to finally settle, forming thus alternating beds of saliferous clay or marl. Such having been the method of formation, it is scarcely necessary THE NONMETALLIC MINEBALS. 197 to state that salt beds are not confined to strata of any one geological horizon, but are to be found wherever suitable circumstances have existed for their formation and preservation. The beds of New York State and of Canada and a part of those of Michigan lie among rocks of the Upper Silurian Age. They are regarded by Professor New- berry as the deposits of a great salt lake that formerly occupied central and western New York, northern Pennsylvania, northeastern Ohio, and southern Ontario, and which he assumes to have been as large as Lake Huron, or possibly Lake Superior. A part of the Michigan beds, on the other hand, were laid down near the base of the Carbon- iferous series, as were also those of the Ohio Valley, and presumably those of Virginia, while those of Petite Anse, Louisiana, are of Cretaceous, or possibly Tertiary Age. The beds of the Western States and Territories are likewise of recent origin, many of them being still in process of formation. The English beds at Cheshire, the source of the so-called "Liver- pool " salt, are of Triassic Age, as are also those of Vic and Dieuze in France, Wurtemburg in Germany, and Salzburg in Austria, while those of Wieliczka in Austrian Poland, and of Parajd in Transylvania are Tertiary. Salt is now manufactured from brines or mined as rock salt in fifteen States of the American Union. These, in the order of their apparent importance, are Michigan, New York, Kansas, California, Louisiana, Illinois, Utah, Ohio, West Virginia, Nevada, Pennsylvania, Virginia, Kentucky, Texas, and Wyoming. At one time Massachusetts was an important producer of salt from sea waters. The industry has, however, been gradually languishing and may ere now be wholly extinct. In California salt is obtained largely from sea water, but also from salt lakes and salines. In Michigan, Ohio, the Virginias, Penn- sylvania, and Kentucky salt is obtained from brines obtained from springs or by sinking wells into the salt-bearing strata, while in New York, Kansas, Louisiana, and the remaining States it is obtained both from brines and by mining as rock salt. Of the foreign sources of rock salt the following districts are the most important: (1) The Carpathian Mountains, (2) the Austrian and Bavarian Alps, (3) western Germany, (4) the Vosges, (5) Jura, (6) Spain, (7) the Pyrenees and the Celtiberian Mountains, and (8) Great Britain, while sea salt is an important product of Turks Island in the Bahamas, of the island of Sicily, and of Cadiz, Spain. We have space here for details concerning but a few of these beds, preference naturally being given to those of the United States. The beds of New York State, of Ontario, northern Pennsylvania, northeastern Ohio, and eastern Michigan all belong to the same geo- logic group are the product of similar agencies. They have been penetrated in many places by wells, and from the results obtained we 198 REPORT OF NATIONAL MUSEUM, 1899. are enabled to form some idea of their extent and thickness. Below is given a summary of results obtained in boring one of these wells to a depth of 1,517 feet at Goderich, Canada. Beginning at the top, the rocks were passed through in the following order: I. Clay, gravel, marls, limestone, dolomite, and gypsum variously interstratified " 7 II. First bed of rock salt 30 n III. Dolomite with marls 32 1 IV. Second bed of rock salt 25 4 V. Dolomite 6 10 VI. Third bed of rock salt 34 10 VII. Marl, dolomite, and anhydrite 80 7 VIII. Fourth bed of rock salt 15 5 IX. Dolomite and anhydrite 7 X. Fifth bed of rock salt 13 6 XI. Marl and anhydrite 135 6 XII. Sixth bed of rock salt 6 XIII. Marl, dolomite and anhydrite 132 Total thickness of formations passed through 1,517 feet. Total thickness of beds of salt 126 feet. The above section shows that the ancient sea or lagoon underwent at least six successive periods of desiccation, and especial attention is called to the remarkable regularity of the deposits. On the oldest sea bottom (XIII) the carbonates and sulphates of lime and magnesia were deposited first, being least soluble. Then followed the salt, and this order is repeated invariably. The other constituents mentioned as occurring in the waters of lakes and seas are not sufficiently abun- dant to show in the section, or owing to their ready solubility they have been in large part removed since the beds were laid down. Chemical tests, however, reveal their presence. Although salt was manufactured from the brine of springs, near Onondaga Lake, in New York, as early as 1788, and has been regu- larly manufactured from the brine of wells since 1798, it was not until subsequent to the discovery of extensive beds of rock salt in the Wyoming Valley, while boring for petroleum, that the mining of the material in this form became an established industry. In June, 1878, a bed of rock salt 70 feet in thickness was found in the valley above mentioned, at a depth of 1,270 feet. Subsequently other borings in Wyoming, Genesee, and Livingston counties disclosed beds at vary- ing depths. In 1885 the first shaft was sunk at Pifford by the Retsof Mining Company, the salt bed being found at a depth of 1,018 feet. Three other shafts have since been sunk, the first about a mile west of the Retsof, the second about 2 miles south of Leroy, and the third at Livonia, in Livingston County. The salt when 'taken from the bed is stated to be of a gray color, due to the presence of clay, which renders solution and recrystallization necessary when designed for culinary purposes. The thickness of the salt beds and their depth are somewhat variable. The following figures are quoted from THE NONMETALLIC MINERALS. 199 Dr. Engelhardt's report. 1 At Morrisville, in Madison County, it is 12 feet thick and at a depth of 1,259 feet; at Tully, in Onondaga County, it varies from 25 to 318 feet, at depths of from 974 to 1,465 feet. The seven beds found at Ithaca have a total thickness of 248 feet, the uppermost lying at a depth of 2,244 feet. In the Genesee Valley the beds vary in depth from 750 to 2,100 feet and in thickness from 40 to 93 feet. In the Wyoming Valley the depth varies from 610 to 2,370 feet below the surface and in thickness from 12 to 85 feet. 2 Michigan. The salt-producing areas of this State are, so far as now known, limited to the counties of losco, Bay, Midland, Gratiot, Saginaw, Huron, St. Clair, Manistee, and Mason, the beds of the Saginaw Valley lying in the so-called Napoleon sandstone, at the base of the Carboniferous. Professor Winchell has estimated this forma- tion to cover an area of some 17,000 square miles within the State limits. The beds of the St. Clair Valley, on the other hand, are in Upper Silurian strata, being presumably continuous with those of Canada. The manufacture of salt from brines procured from these beds began in the Saginaw Valley in 1860 and has since extended to the other regions mentioned. According to F. E. Engelhardt the rock salt deposits in the Upper Silurian beds, with a thickness of 115 feet, were reached at Marine City, in St. Clair County, at a depth of 1,633 feet; at St. Clair, St. Clair County, at a depth of 1,635 feet and with a thickness of 35 feet. At Caseville, in Huron County, the beds lie at a depth of 1,164 feet, and at Bay City, Saginaw Bay, at 2,085 feet, the salt beds being 115 feet in thickness. At Manistee the bed is 34 feet thick, lying 2,000 feet below the surface, while at Muskegon, in the Mason well, it was 50 feet thick at a depth of 2,200 feet. Although of so recent development, Michigan is rapidly becoming one of the leading salt-producing regions of the world, the estimated manu- facturing capacity being now upward of 5,000,000 barrels annually. The total product of all the years since 1868 is given as 60,614,464 barrels of 280 pounds each. In Kansas the rock salt occurs in beds regarded as of Permian age, and has been reached by means of shafts in several counties in the southern and central part of the State. The following is a section of a shaft sunk at Kingman in 1888-89: Feet. "Red-beds," red arenaceous, limestones, ferruginous clays, and clay shales with thin streaks of gray shales and bands of gypsum as satin spar 450 Gray or bluish ' < slate, ' ' with 2 feet of limestone at 500 feet 140 Red clay shale 4 Gray "slate," with occasional streaks of limestone, 2 to 8 inches thick, and some salt partings and satin spar with ferruginous stain 78 1 The Mineral Industry, its Statistics and Trade for 1892, by R. P. Rothwell. 2 For a very complete historical and geological account of these salt beds and the method of manufacture, see Bulletin No. 11, of the New York State Museum, 1893, by F. J. H. Merrill. 200 BEPOBT OF NATIONAL MUSEUM, Feet. 2 First rock salt, pure white ............................... '~'~~" f ---- ~"~" Shale and "slate," bluish, with vertical and other seams of salt, from 1 to 3 inches thick ............................................................ 26 Rock salt ........................... ..................................... 4 Shales, with salt ......................................................... n Kocksalt ................................................................ 7 Shale .................................................................... 3 Rock salt ................................................................ Salt and shale, alternate thin seams ......................... - ............. < Rock salt ........................................ Shale ........................... l Rock salt ................................................................ 5 Shales and limestone ..................................................... Rock salt, bottom of it not reached ......................... - ............. 5 Total ....................................... -- 820 Borings and shafts have also proven the existence of beds of salt in other parts of the State, as at Kanopolis, Lyons, Caldwell, Rago, Pratt, and Wilson. According to Dr. Robert Hays l it is safe to assume that beds of rock salt from 50 to 150 feet in thickness underlie fully half the area from the south line of the State to north of the Smoky River, an area from 20 to 50 miles in width. Although the mining of rock salt began in this region only in 1888, the annual output has already reached over 1,000,000 barrels. ^Louisiana. Salt in this State is derived from Petite Anse, a small island rising from the marshes on the southern coast and connected with the mainland by a causeway some 2 miles in length. According to E. W. Hilgard 2 the deposit is probably of Cretaceous age, and is presumably but a comparatively small residual mass of beds once extending over a much larger area, but now lost through erosion. (See fig. 3.) Exploration has shown the area occupied by the beds to be some 150 acres, but the full thickness, though known to be upward of 165 feet, has never been fully determined. Salt in Kentucky is obtained from the brine of springs and wells in Carboniferous limestone. In Meade County brine accompanies the natural gas, the latter in some cases being utilized as fuel for its evap- oration. Springs in Webster County furnished salt for Indians long anterior to the occupancy of the county by whites, and fragments of their clay kettles and other utensils used in the work of evaporation are still occasionally found. Texas. The occurrences of salt are numerous and widespread. Along the coast are many lagoons and salt lakes, from which considerable quantities are taken annually. "Besides the lakes along the shores many others occur through western Texas, reaching to the New Mexico 'Geological and Mineral Resources of Kansas, 1893, p. 44. "Smithsonian Contributions to Knowledge, XXIII. Qn the Geology of Lower Louisiana and the Salt Deposit on Petite Anse Island. THE HONMETALLIC MINERALS. 201 w 3 II u b . s ^ : : 202 REPOBT OF NATIOKAL MUSEUM, 1899. line while northeast of these, in the Permian region the constant recurrence of such names as Salt Fork, Salt Creek, etc., tell of the prevalence of similar conditions." In addition to the brines there are extensive beds of rock salt. That which is at present best developed is located in the vicinity of Colorado City, in Mitchell County. The bed of salt was found at a depth of 850 feet, with a thickness ot 140 feet In eastern Texas there are many low pieces of ground calle salines, where salt has been manufactured by evaporation of the brines obtained from shallow wells. At the "Grand Saline,' in Van Zandt County, a bed of rock salt over 300 feet in thickness was found at a depth of 225 feet. In England the salt occurs at Cheshire in two beds mterstratified with marls and clays. The upper, with a thickness varying from 80 to 90 feet, lies at a depth of some 120 feet below the surface, and the second at a depth of 226 feet has a thickness varying between 96 and 117 feet. The accompanying general sections are from Davies' Earthy and Other Economic Minerals. Detailed section of strata sunk through at Witton, near Northwich, to the lower bed of suit. Ft. In. 1. Calcareous marl 2. Indurated red clay 3. Indurated blue clay and marl 4. Argillaceous marl 5. Indurated blue clay 6. Red clay with sulphate of lime in irregular branches. . . 7. Indurated red clay with grains of sulphate of lime interspersed -4 8. Indurated brown clay with sulphate of lime crystallized in irregular masses and in large proportions 9. Indurated blue clay with laminae of sulphate of lime 10. Argillaceous marl. . .- 11. Indurated brown clay laminated with sulphate of lime 3 12. Indurated blue clay laminated with sulphate of lime 3 13. Indurated red and blue clay 12 14. Indurated brown clay with sand and sulphate of lime irregularly inter- spersed through it. The fresh water, at the rate of 360 gallons a minute, forced its way through this stratum 13 15. Argillaceous marl 5 16. Indurated blue clay with sand and grains of sulphate of lime 3 9 17. Indurated brown clay as next above 15 18. Blue clay as strata next above ' : 1 6 19. Brown clay as strata next above j 7 20. The top bed of rock salt 75 21. Layers of indurated clay with veins of rock salt running through them. . .31 6 22. Lower bed of rock salt... ..115 Total 341 9 At Wieliczka, in Austrian Poland, the salt occurs in massive beds stated to extend over an area some 20 by 500 miles, with a maximum thickness of 1,200 feet. At Parajd, in Transylvania, beds belonging THE NONMETALLIC MINERALS. 203 to the same geological horizon are estimated to contain upward of 10,000,000,000,000 cubic feet of salt. One of the most remarkable deposits of the world, remarkable for its extent as well as for the variety of its products, is that of Stass- furt, in Prussian Saxony. On account of its unique character, as Fig. 4. CLUSTER OF 8YLVITE CRYSTALS. Stassfurt, Germany. Specimen No. 10223, U.S.N.M. well as its commercial importance, being to-day the chief source of natural potash salts of the world, a little space may well be given here to a detailed description. 1 Stassfurt is a small town of some 12, 000 inhabitants, about 25 miles southwest of the city and fortress of Magdeburg, in Prussia. It lies in a plain, and the river Bode, which takes its rise in the Harz Mountains, flows through it. The history of the salt industry in Stassfurt is a very old one, and dates back as far as the year 806. Previous to the year 1839 the salt was produced from brine pumped from wells sunk about 200 feet into the rock. The brine, in the course of time, became so weak, Journal of the Society of Chemical Industry, II, 1883, pp. 146, 147. 204 BEPOKT OF NATIONAL MUSEUM, 1899. as regards the common salt it contained, that it was impossible to carry on the manufacture from this source without loss. In 1839 the Prussian Government who were the owners of these saline springs, commenced boring with the object of dis- covering the whereabouts of the bed of rock salt from which the brine had been obtained and in the year 1843, seven years after the commencement of the borings, the top of the rock salt was reached at a depth of 256 metres. The boring was con- tinued through another 325 metres into the rock salt without reaching the bottom of the layer. At this total depth of 581 metres the boring was suspended. On ana- lysing the brine obtained from the bore-hole, it was found to consist, in 100 parts by weight, of- Sulphateofcalcimn 4.01 Chloride of potassium 2. 24 Chloride of magnesium - 19. 43 Chloride of sodium 5. 61 A result not only unexpected but disappointing, since the presence of chloride of magnesium in such quantities dispelled for the time all hopes of striking on the pure rock salt. The Government, however, guided by the opinions expressed by Dr. Karsten and Professor Marchand, namely, that the presence of chloride of magnesium in such quantities was probably due to a deposit lying above the rock salt, determined to further investigate the matter, and in the year 1852 the first shaft was commenced, which after five years had penetrated, at a depth of 330 metres, into a bed of rock salt, passing on its way, at a depth of 256 metres, a bed of potash and magnesia salts of a thickness of 25 metres. On referring to the section of the mines [Plate 4], it will be seen that the lowest deposit of all consists of rock salt. The bore-hole was driven 381 metres into it without reaching the bottom of the layer. Its depth is therefore unknown. The black lines drawn across the rock salt deposit represent thin layers of sulphate of cal- cium 7 millimetres thick, and almost equidistant. The lines at the top of the rock salt represent thin layers of the trisulphate of potash, magnesia, and lime as the mineral Polyhallite [Specimen No. 67754, U.S.N.M.]. The deposit lying immediately on the bed of rock salt consists chiefly of sulphate of magnesia as the mineral Kie- serite [Specimen No. 62417, U.S.N.M.]. Still farther toward the surface the deposit consists of the double chloride of potassium and magnesium, known as the mineral Carnallite, [Specimens Nos. 40225, 62416, U.S.N.M.] mixed with sulphate of magnesia and rock salt. The deposit to the right, on the rise of the strata, consists of the double sulphate of potash and magnesia combined with one equivalent of chloride of magnesium, and intermingled with common salt to the extent of 40 per cent. This double sulphate is known as the mineral Kainite [Specimen No. 64735, U. S. N. M. ] and is a secondary formation, resulting from the action of a limited quantity of water on a mixture of sulphate of magnesia and the double chloride of potassium and magnesium, as contained in the uppermost deposit previously spoken of. The upper bed of the rock salt, resting on a thick bank of Anhydrite [Specimen No. 64740, U.S.N.M.], is also a later formation. Almost imperceptible layers of Poly- hallite are present in this deposit and at greater intervals than in the lower and older deposit. It has therefore probably originated from the action of water on the older deposit. This upper bed of rock salt varies in thickness from 40 to 90 metres, and its extent is comparatively limited. It is worked in preference to the older deposit, where both exist in the same mine, it being of much purer quality, aver- aging about 98 per cent in the mines of the New Stassfurt Mining Company and in the Royal Prussian mines. Sixteen different minerals have as yet been discovered in the Stassfurt deposits. They may be divided into primary and secondary formations. Those of primary formation are rock salt, Anhydrite [Specimen No. 64740, U.S.N.M.], Polyhallite (K 2 S0 4 , MgS0 4) 2CaS0 4 , 2H 2 0) [Specimen No. 67754, U.S.N.M.], Kieserite (MgSO 4 , Report of U. S. National Museum, 1899. Merrill. Prussian 5 hafts. PLATE 4. ., x ^(> Sen wabire, Hall Stassfurt Hallstadt in Up. Aus- tria. Wilhelmsgliick 99.63 94.57 98.14 98 36 Tr. 0.09 0.28 0.97 0.89 1 86 .... 1.12 2. 'S.', >.22 ii. 56 ) .">(! 0.03 1. 1;:> 1.58 ) "0 Vic in German Lo- raine. Jeb-el-Melah Algeria 99.30 ; IK) Ouled Kebbah.Algeria Cheshire, England Carrickfergus, Ireland. Holston Virginia 98.53 99.32 96.28 99 55 0.93 0.57 0.02 ). -1C, 1.50 0.08 i i") Ul Tr Petite Anse, Louisiana. Santo Domingo 98.88 98.33 98.55 96.76 97.21 99.77 99.85 Tr. 0.99 Tr. 0.04 0.02 0.14 0.26 0.01 0.03 i. 71) 1.48 i 1 1 i ;; 1.01 1.07 Cardona, Spain Sea sail. Turks Island St. Martins StKitts Curacoa 1.66 1.54 0.08 I' 1 0.64 0.24 I. '.K,) 1 . 75 ) 1 1 Cadiz 95.76 94.17 96.78 94.91 99.46 98.435 96.36 97.39 96.93 97.03 97.41 96.70 91.31 99.11 92.97 93.07 96.42 98.12 9*. 06 0.57 1.11 0.49 ?4 .... 0.49 1.43 1 4? 0.48 1.39 0.68 0.19 ) 11 Lisbon ... ' SJ Trapani, Sicily Marthas Vineyard Texas Pacific coast (Union Pacific Salt Co.). Salt from springs and Cheshire, England Dienze, German Lo- raine. Droitwich, England... Goderich, Ontario Onondaga, New York . . Pittsburg, Pennsylva- nia. Kanawha, West Vir- ginia. Holston, Virginia Saginaw, Michigan.... Hocking Valley, Ohio. Pomeroy, Ohio Nebraska --- 1.64 3 24 365 1 '"0 0.01 0.02 1 17 0.89 J. f.o 0.02 ; o: 0.01 0.15 0.33 1.26 0.03 0.18 0.07 0.43 1 r> i.2< 1 (K) > 7() 1.09 0.61 0.53 0.50 0.04 0.18 07 0.68 i. :;; I. H 0.11 .... o. or, 0. 01 0.16 0.10 ').!() !. -W j. c>t; .). SO 1.80 Kansas 0.241.... 1.12 0.18 THE NONMETALLIC MINERALS. 213 Composition of salt from various localities Continued. Varieties of salt. Chloride of sodium. i. i 1 "2, 3 Chloride of cal- cium. Chloride of magne- sium. Sulphate of potash. Sulphates of cal- cium. Sulphates of magne- sia and soda. Carbonates of mag- nesia and lime. Alumina and iron. I Water. Percentage of sa- line residue. Authorities. Sail from springs and lakes. Cont'd. Onondaga "factory filled." Great Salt Lake 98.28 97 61 0.91 LOS 0. 51 > o-i 0.09 0.08 0.35 .... ).12 ). (10 1 '>S Goessman. G.H.Cook. Gobel. Meissner. Heine. Herman. Watts Diet, of Chem., Vol.V, p. 334. Heine. Do. Bromeis. Figuer and Mi- alho. Wm. Henry. G. H. Cook. Boussingault. G.H.Cook. Do. Do. Do. Do. Usiglio. Rose. Booth and Muckle. L. D. Gale. F. Gutzkow. Elton Lake, Russia Solid residue of brines and sea water. Halle, in Prussia and Saxony. Stassfurt 98.95 94.43 94.49 95.71 93.72 95.35 89.88 0.21 1.03 0.19 1.69 0.99 1.09 0.67 1.59 i -,' 12.28 17.16 2.00 11.10 26.50 8.39 2.87 1.27 26.00 15.20 21.20 18.54 80 ).;;i >. OS 1.34 1 1i 2.80 l.iil 2.55 i :>i 1.20 1.37 1.18 I. IS i.or, ). 19 0. (>: Tr. .... Schonebeck 0.08 >. 15 Do Artern, from bore in rock salt. 1.49 1.18 2.24 0.25 >. 99 0.04 US 1.66 1 on 0.63 >. 17 7.63 s. 79 i. 77 i ( ).()_ 0.07 1.51 Manheim Soden 82.23 86.01 97.40 84.87 1.88 1.81 6.74 0.25 Cheshire Dieuze 1.83 :.o'. 3.30 China 75.47 95 42 -... 17.92 0.84 13.93 16.48 5.97 0.64 4.80 4.07 Mil Tr Pittsburg Kanawha Holston 81.27 79.45 98 39 TV 9.20 26.40 24 90 ] .,.> 39 Tr Salt Lake Texas 97 08 i. 82 ;. 17 1.87 2.10 6.42 18.26 8.18 Sea water 78.61 13.15 29.86 90.07 1.34 ). 79 2.51 11.81 8.56 67.80 55.45 1.12 .... 0. l'7 3.74 29.13 26.42 22.42 3.038 Elton Lake Dead Sea Great Salt Lake Sea water (San Fran- Cisco Bay). 2. FLUORITE. This is a calcium fluoride, CaF 2 = fluorine 48.9 per cent, cal- cium 51.1 per cent. The most striking features of this mineral are its cubic crystallization (Specimens Nos. 51226, 66831, 66832, U.S.N.M.), octahedral cleavage (Specimen No. 48270, U.S.N.M.), and fine green (Specimen No. 48270, U.S.N.M.), yellow (Specimen No. 49160, U.S.N.M.), purple (Specimen No. 51226, U.S. N.M.), violet, and sky blue colors. White (Specimen No. 36091, U.S.N.M.) and red- 214 REPOKT OF NATIONAL MUSEUM, 1899. brown varieties are also known. The mineral is translucent to trans- parent, and of a hardness somewhat greater than calcite (4 of Dana's scale). Occurrence. The mineral occurs as a rule in veins, though some- times in beds in gneiss, the schists, limestones, and sandstones. It is also a common gangue of metallic ores, particularly those of lead and tin. At Rosiclare, in southern Illinois, the fluorspar veins, according to Emmons, 1 are true fissure veins, varying from 4 to 20 feet in width in limestones immediately underlying the coal measures. He regards the original crevice as formed by dynamic action, as probably com- paratively small and subsequently enlarged by solution by percolat- ing waters. The source of the fluorspar of the veins would seem to be the surrounding limestones. The associated minerals are galena and calcite, with smaller quanti- ties of sphalerite and iron and copper pyrites. Uses. The material is used mainly as a flux for iron, in the manu- facture of opalescent glass and for the production of hydrofluoric acid. The chief source of supply in the United States is Rosiclare, Illinois, the annual output being some 6,000 to 10,000 tons, valued at about $5 a ton. 3. CRYOLITE. Composition. Na 3 AlF 6 ,= aluminum 12.8 per cent; sodium 32.8 per cent; fluorine 54.4 per cent. The mineral is as a rule of snow-white color, though sometimes reddish or brownish, rarely black, and coarsely crystalline granular, translucent to subtransparent. It has a hardness of 2.5; specific gravity of 2.9 to 3, and in thin splinters may be melted in the flame of a candle. The name is from the Greek word /c-p^o?, ice, in allusion to its trans- lucency and ice-like appearance (Specimen No. 17571, U.S.N.M.). Mode of occurrence. Cryolite occurs, as a secondary product, in the form of veins. It is rarely found in sufficient abundance to be of commercial value, the supply at present coming almost wholly from Evigtok in South Greenland. The country rock here is said to be granite and the vein as described in 1866 2 was 150 feet in greatest breadth and was exposed for a distance of 600 feet. The principal mineral of the vein was cryolite, but quartz, siderite, galena, and chal- copyrite were constant accompaniments, irregularly distributed through the mass. In 1890 the mine as worked was described as a hole in the ground elliptical in shape, 450 feet long by 150 feet wide, the pit being some 100 feet deep. The drills had penetrated 150 feet transactions of the American Institute of Mining Engineers, XXI, 1893, p. 31. 2 Paul Quale, Report of Smithsonian Institution, 1866, p. 398. . THE NONMETALLIC MINERALS. 215 deeper and found cryolite all the way. Johnstrup, as quoted by Dana, 1 describes the cryolite as: Limited to the granite; he distinguishes a central and a peripheral part; the former has an extent of 500 feet in length and 1,000 feet in breadth and consists of cryolite chiefly, with quartz, siderite, galena, sphalerite, pyrite, chalcopyrite, and wolframite irregularly scattered through it. The peripheral portion forms a zone about the cen- tral mass of cryolite; the chief minerals are quartz, feldspar, and ivigtite, also fluor- ite, cassiterite, molybdenite, arsenopyrite, columbite. Its inner limit is rather sharply denned, though there intervenes a breccia-like portion consisting of the minerals of the outer zone enclosed in cryolite; beyond this it passes into the surrounding granite without distinct boundary. Cryolite in limited quantity occurs at the southern base of Pike's Peak, in Colorado, and north and west of St. Peter's Dome (Specimen No. 48220, U.S.N.M.). It is found in vein-like masses of quartz and microcline embedded in granite. Uses. Until within a few years the material has been utilized only in the manufacture of soda, and sodium and aluminum salts, and to a small extent in the manufacture of glass and porcelain ware. It is also used in the electrotytic processes of extracting aluminum from its ores, as now practiced. The principal works utilizing the Greenland cryolite in chemical manufacture are, at time of writing, those of the Pennsylvania Salt Manufacturing Company at Natrona, Pennsylvania (see series of crude and manufactured products Nos. 6332T to 63334, TJ.S.N.M). IV. OXIDES. 1. SILICA. QUARTZ. The mineral quartz, easily recognized by its insolubility in acids, glassy appearance (Specimen No. 67985, U.S.N.M.), lack of cleavage, and hardness, which is such that it readily scratches glass, is one of the most common and widely disseminated of minerals. Chemically it is pure silica, of the formula SiO 2 . It crystallizes in the hexagonal system with beautiful terminations, and is one of the most attractive of minerals for the amateur collector (Specimen No. 61768, U.S.N.M.). The common form is, however, massive, occurring in veins in the older crystalline rocks (Specimen No. 55244, U.S.N.M.). Common sand is usually composed mainly of quartzose grains which, owing to their hardness and resistance to atmospheric chemical agen- cies, have withstood disintegration to the very last. The terms rose, milky (Specimen No. 62381, U.S.N.M.), and smoky (Specimen No. 67986, TJ.S.N.M.) are applied to quartzes which differ from the ordinary type only in tint, as indicated. Chalcedony is the name given to a somewhat hornlike, translucent or transparent form of silica occurring only as a secondary constituent in veins, or isolated con- cretionary masses, and in cavities in other rocks. Agate is a banded System of Mineralogy, 1892, p. 167. 216 REPORT OF NATIONAL MUSEUM, 1899. variety of chalcedony. The true onyx is similar to agate, except that the bands or layers of different colors lie in even planes. Jasper is a ferruginous, opaque chalcedony, sometimes used for ornamental pur- poses. Opal is an amorphous form of silica, containing somewhat variable amounts of water. Quartz occurs as an essential constituent of granite, gneiss, mica schist, quartz porphyry, and liparite, and also as a secondary constitu- ent in' the form of veins, filling joints and cavities in rocks of all kinds and all ages. fos._The finer clear grades of quartz are used to some extent for spectacle lenses and optical work, as well as in cheap jewelry (Specimen No. 11893, U.S.N.M.). Its main value is, however, for abrading pur- poses, either as quartz sand or as sandpaper (Series Nos. 55877-55884, U.S.N.M.), and in the manufacture of pottery (Specimens Nos. 62123, 63035-63038, U.S.N.M.). For abrading purposes it is crushed and bolted, like emery and corundum, and brings a price barely sufficient to cover cost of handling and transportation. Pure quartz sand is also of value for glass making (Specimens Nos. 53188, 60683, 63128, 63123, 63122, U.S.N.M.), and ground quartz to some extent as a "filler" in paints (Specimen No. 63119, U.S.N.M.), and as a scouring material in soaps. The following analyses show the composition of some glass sands from (I) Clearfield and (II) Lewistown, Pennsylvania: Constituents. I. II. Silica 99.79 98.84 0.12 0.17 Iron oxides 0.014 0.34 Lime 8 Traces. Ignition 23 100. 724 99.58 FLINT is a chalcedonic variety of silica found in irregular nodular forms in beds of Cretaceous chalk. These nodules break with a con- choidal fracture and interiorly are brownish to black in color (Speci- men No, 62120, U.S.N.M.). By the aboriginal races the flints were utilized for the manufacture of knives and general cutting imple- ments. Later they were used in the manufacture of gun flints and the "flint and steel" for producing fire. At present they are used to some extent in the manufacture of porcelain, being calcined (Speci- men No. 62061, U.S.N.M.) andground (Specimen No. 62122, U.S.N.M.) to mix with the clay and give body to the ware. In this country the same purpose is accomplished by the use of quartz. Small round nodules of flint from Dieppe, France, are said to be used in the Tren- ton (New Jersey) pottery works for grinding clay by being placed in revolving vats of water and kaolin. All the flint now used in this country is imported either as ballast or as an accidental constituent of chalk. THE NONMETALLIC MINEBALS. 217 As the material is worth but from $1 to $2 a ton delivered at Trenton, it may be readily understood that transportation is a rather serious item to be considered in developing home resources. According to Mr. R. T. Hill, nodules of black flint occur in enormous quantities in the chalky limestones the Caprina limestones of Texas. Numerous localities are mentioned, the most accessible being near Austin, on the banks of the Colorado River. BUHRSTONE, or burrstone, is the name given to a variety of chalcedonic silica, quite cavernous, and of a white to gray or slightly yellowish color. The cavernous structure is frequently due to the dissolving out of calcareous fossils. The rock is of chemical origin that is, results from the precipitation of silica from solution, and pre- sumably through the action of organic matter. In France the material occurs alternating with other unaltered Tertiary strata in the Paris basin (Specimen No. 36140, U.S.N.M.). It is also reported in Eocene strata in South America, and in Burke and Screven counties along the Savannah River in southern Georgia in the United States (Specimen No. 36051, U.S.N.M.). The toughness of the rock, together with the numerous cavities, impart a sharp cutting power such as renders them admirably adapted for millstones, and in years past material for this purpose has been sent out from French sources all over the civilized world. TRIPOLI is the commercial name given to a peculiar porous rock regarded as a decomposed chert associated with the Lower Car- boniferous limestones of southwest Missouri (Specimen No. 55028, U.S.N.M.). The rock is of a white cream or slight pink cast, fine grained and homogeneous, with a distinct gritty feel, and, though soft, sufficiently tenacious to permit of its being used in the form of thin disks of considerable size for filtering purposes (Specimen No. 62044, U.S.N.M.). According to Hovey 1 the deposit is known to underlie between 80 and 100 acres of land, in the form of a rude ellipse, with its longer diameter approximately north and south. From numerous prospect holes and borings it has been shown to have an average thick- ness of 15 feet, the main quarry of the present company showing a thickness of 8 feet. The following section is given from a well sunk in the northern part of the area: Feet. Earth to 4 Tripoli 4 20 Stiff red clay 20 21% Mixed chert, clay, and ochre 21 40 Cherty limestone 40 93 Cherty limestone bearing galena 93 103 Limestone 103 128 Limestone bearing sphalerite and galena 128 136 Soft magnesian limestone 136 173 1 Scientific American Supplement, July 28, 1894, p. 15487. 218 REPORT OF NATIONAL MUSEUM, 1899. The tripoli is everywhere underlain by a relatively thin bed of stiff red clay, and also traversed in every direction by seams of the same material from 1 to 2 inches thick. These seams and other joints divide the rock into masses which vary in size up to 30 inches or more in diameter. Microscopic examinations as given by Hovey show the rock to contain .no traces of organic remains, but to be made up of faintly doubly refracting chalcedonic particles from 0.01 to 0.03 milli- metre in diameter. The chemical composition, as shown from analysis by Prof. W. H. Seaman, is as follows: Silica (Si0 2 ) 98.100 Alumina (A1 2 3 ) 0-240 Iron oxide (Fe O and Fe 2 O s ) 0. 270 Lime (CaO) 0.184 Soda (Na,O) 0.230 Water (ignition) 1. 160 Organic matter 0. 008 100. 192 Silica soluble in a 10 per cent solution of caustic soda on boiling three hours, 7.28 per cent. Aside from its use as a filter (Specimens Nos. 62044 and 62045, U.S.N.M.) the rock is crushed between burr stones, bolted, and used as a polishing powder (Specimens Nos. 51231 and 55029, U.S.N.M.). To a small extent it has been used in the form of thin slabs for blotting purposes, for which it answers admirably owing to its high absorptive property, but is somewhat objectionable on account of its dusty char- acter. The view (Plate 6) shows the character of a quarry of this material as now worked by the American Tripoli Company at Seneca, in Newton County. DIATOMACEOUS OR INFUSORIAL EARTH, as it is sometimes called, is, when pure, a soft, pulverulent material, somewhat resembling chalk or kaolin in its physical properties, and of a white or yellow- ish or gray color. Chemically it is a variety of opal (see analyses on page 220). Origin and occurrence of deposits. Certain aquatic forms of plant life known as diatoms, which are of microscopic dimensions only, have the power of secreting silica, in the same manner as mollusks secrete carbonate of lime, forming thus their tests or shells. On the death of the plant the siliceous tests are left to accumulate on the bottom of the lakes, ponds, and pools in which they lived, forming in time beds of very considerable thickness, which, however, when compared with other rocks of the earth's crust are really of very insignificant propor- tions. Like many other low organisms the diatoms can adapt them- selves to a wide range of conditions. They are wholly aquatic, but live in salt and fresh water and under widely varying conditions of Report of U. S. National Museum, 1899. Me PLATE 6. Report of U. S. National Museum, 1899. Me PLATE 7. DEPOSIT OF DIATOMACEOUS EARTH, GREAT BEND OF PITT RIVER, SHASTA COUNTY, CALIFORNIA. From fi photograph by J. S. Diller. THE NONMETALLIC MINERALS. 219 depth and temperature. They may be found in living forms in almost any body of comparatively quiet water in the United States. The exploring steamer Challenger dredged them up in the Atlantic from depths varying from 1,260 to 1,975 fathoms and from latitudes well toward the Antarctic Circle. Mr. Walter Weed, of the U. S. Geolog- ical Survey, has recently reported them as living in abundance in the warm marshes of the Yellowstone National Park, while Dr. Blake reported finding over 50 species in a spring in the Pueblo Valley, Nevada, which showed a temperature of J 63 F. Although beds of diatomaceous earth are still in process of forma- tion, and in times past have been formed at various epochs, the Tertiary period appears for some reason to have been peculiarly fitted for the growth of these organisms, and all of the known beds of any impor- tance, both in America and foreign countries, are of Tertiary age. The best known of the foreign deposits is that of Bilin, in Bohemia. This is some 14 feet in thickness. When it is borne in mind that, according to the calculations of Ehrenberg, every cubic inch of this contains not less than 40,000,000 independent shells, one stands aghast at the mere thought of the myriads of these little forms which such a bed repre- sents. Some of the deposits in the United States are, however, con- siderably larger than this. What is commonly known as the Richmond bed extends from Herring Bay, on the Chesapeake, Maryland, to Petersburg, Virginia, and perhaps beyond. This is in some places not less than 30 feet thick in thickness, though very impure (Specimen No. 67984, U.S.N.M., from Calvert County, Maryland, is fairly repre- sentative). Near Drakes ville, in New Jersey, there occurs a smaller deposit, covering only some 3 acres of territory to a depth of from 1 to 3 feet. Some of the largest deposits known are in the West. Near Socorro, in New Mexico, there is stated to be a deposit of fine quality which crops out in a single section for a distance of 1,500 feet and some 6 feet in thickness. Geologists of the fortieth parallel survey reported abundant deposits in Nevada, one of which showed in the railroad cutting west of Reno a thickness not less than 300 feet, and of a pure white, pale buif, or canary yellow color (Specimen No. 67916, U.S.N.M.). Along the Pitt River, in California, there is stated to be a bed extending not less than 16 miles and in some places over 300 feet thick (see Plate 7). Near Linkville, Klamath County, Oregon (Specimens Nos. 53402, 53093, U.S.N.M.), there occurs a deposit which has been traced for a dis- tance of 10 miles, and shows along the Lost River a thickness of 40 feet. Beds are known also to occur in Idaho (Specimens Nos. 63843, 66950, U.S.N.M.), near Seattle, in Washington (Specimen No. 53200, U.S.N.M.), and doubtless many more yet remain to be discovered. A deposit of unknown extent, pure white color, and almost pulp-like consistency has been worked in South Beddington, Maine (Specimens 220 EEPOBT OF NATIONAL MUSEUM, 1899. Nos. 73253, 73254, U.S.N.M.). Others of less purity occur near South Framingham, Massachusetts (Specimens Nos. 62767, 62768, U.S.N.M.), Lake Umbagog, New Hampshire (Specimen No. 29322, ILS.N.M.), at White Head Lake, Herkimer County, New York (Spec- imen No. 62913, U.S.N.M.), and at Grand Manan, New Brunswick (Specimen No. 57339, U.S.N.M.). Chemical Composition. As already intimated, this earth is of a siliceous nature, and samples from widely separated localities show remarkable uniformity in composition. Of the following analyses, No. 1 is from Lake Umbagog, New Hampshire, No. II, from Morris County, New Jersey, and No. HI, from Popes Creek, in Maryland. As will be noted, the silica percentage is nearly the same in all. Constituents. I II III Silica 80.53 80.66 81.53 5.89 3.84 3.43 Iron oxides Lime 1.03 0.35 0.58 3.34 2.61 Soda 1 43 Potash 1.16 12 03 14 01 6 04 The substance may therefore be regarded as a variety of opal. Uses. The main use of infusorial earth is for a polishing powder. It is, however, an excellent absorbent, and has been utilized to mix with nitroglycerine in the manufacture of dynamite. It has also been used to some extent in the preparation of the soluble silicate known as water glass. The demand for the material is therefore quite smal 1 , not nearly equal to the supply. The Maryland and Nevada deposits are said to be the principal ones now worked. During the year 1897 the entire output was about 3,000 tons, valued at some $30,400. 2. CORUNDUM AND EMERY. CORUNDUM. Composition, sesquioxide of aluminum A1 2 O 3 , = oxygen, 47.1 per cent; aluminum, 52.9 per cent. In crystals often quite pure, but frequently occurring associated in crystalline granular masses with magnetic iron, and often more or less altered into a series of hydrated aluminous compounds, as darnourite (Specimen No. 82492, U.S.N.M.). The crystalline form of the mineral is hexagonal, or sixsided in out- line, and often with curved sides and square terminations, giving rise to roughly barrel-shaped forms, as shown in specimen No. 81450 from Bengal, India. A prominent basal cleavage causes the crystals to break readily with smooth, flat surfaces at right angles with the axis of elongation. The massive forms often show a nearly rectangular parting or pseudo- THE NONMETALLIC MINERALS. 221 cleavage (Specimen No. 63480, U.S.N.M., from Pine Mountain, Georgia). The most striking physical property of the mineral is its hardness, which is 9 of Dana's scale. In this respect it ranks then next to the diamond. The color of the mineral varies from white through gray (Specimen No. 46283, U.S.N.M), brown, yellow, blue (Specimens Nos. 73531 and 48182, U.S.N.M.), pink (Specimen No. 81922, U.S.N.M.), and red; luster adamantine to vitreous; specific gravity, 3.95 to 4.1. The highly colored transparent red and blue forms are valuable as gems, and are known under the names of ruby and sapphire. The consideration of these forms is beyond the limits of this work. (See Mineral and Gem Collections.) Occurrences. Athough widespread as a mineral, corundum, unmixed with a large proportion of magnetite (forming emery), has been found in but few localities in sufficient abundance to be of commercial value. The most important deposits in the United States are in southwestern North Carolina and in the Laurel Creek region of northern Georgia. The country rock in both these regions is hornblendic gneiss, through which has been intruded a basic eruptive (dunite, Specimen No. 70069, U.S.N.M.), and it is mainly along the decomposed lines of contact between the two that the corundum is found. According to Dr. T. M. Chatard, the Corundum Hill Mine is situated on a ridge which runs in the northeast and southwest direction characteristic of this section, the dunite outcrops being on the crest, and apparently surrounded on all sides except toward the east by hornblende gneiss. On the east side mica schist (probably damourite schist) takes the place of the gneiss, and it is on the eastern side of the dunite that the so-called " sand vein" is found. This is a vein-like mass of brown vermiculite in small scales containing an abundance of small crystals of corundum which are usually brown in color and often broken into fragments (Specimen No. 73529, U.S.N.M.). The easterly wall of this vein is the mica schist very much decomposed, while on the western side is found enstatite (Speci- men No. 70070, U.S.N.M.), next vermiculite mixed with chlorite, then talc (Specimen No. 70071, U.S.N.M.), which in turn gives place to nodules of more or less altered dunite. The specimens of corundum crystals for which this locality is so celebrated (Specimen No. 73530, U.S.N.M.) have been found mainly, if not wholly, on the westerly side of the dunite, and on or near the line of contact between the gneiss and dunite. State Geologist Yeates has stated 1 that in the Laurel Creek region the corundum is not confined to the vermiculite and chlorite bands, but is abundant in the lime soda feldspar as well. The same authority states that in this region the dunite is not inclosed by the hornblendic Bulletin No. 2, Geological Survey of Georgia, 1894. 222 REPORT OF NATIONAL MUSEUM, 1899. gneisses, but intruded between these and other gneiss or mica schist; also that the corundum-bearing veins lie in the dunite close to the con- tact and in the vicinity of the hornblendic gneiss. It should be said before leaving the subject that certain micaceous minerals, as margarite and chloritoid (Specimen No. 63107, U.S. KM., from Chester, Massa- chusetts) are almost invariable accompaniments of corundum and emery deposits, and that it was the finding of these minerals that led to the discovery of the emery beds at Chester. Chatard reports that in the North Carolina mines chlorite or vermiculite is considered a "corundum sign," and in mining such indications are followed so long as they hold out (Specimen No. 63153, U.S.N.M.). The geographical distribution of corundum-bearing rocks in the eastern United States has been worked out in detail by J. V. Lewis of the North Carolina Geological Survey, from whose report 1 the accom panying map (Plate 8) is taken. According to this authority the corundum occurring in such quantities as to be of commercial value is almost universally found in connection with basic eruptive rocks, as peridotites or their varietal forms pyroxenite and amphibolite, which are themselves intruded into gneisses. At Yogo Gulch, Montana, corundum in the form of sapphire (see Gem Collections) occurs as a constituent of a basic eruptive rock near the line of contact with aluminous shales (Specimen No. 53519, U.S.N.M.). In Gallatin County the mineral is found in well-defined crystals of all sizes up to an inch or more in length abundantty disseminated through- out a granite (Specimen No. 83838,U.S.N.M.). In the Russian Urals it occurs in disseminated crystals and large cleavage masses in feldspar (Specimens Nos. 40323, 40315, 40334, 73532, U.S.N.M.). In India it occurs as an original constituent associated with both acid and basic rocks, but in most cases where the mineral is in the basic rocks there have been found intrusions of pegmatite (an acid rock) in the near vicinity. In the celebrated Mogok Ruby Mines the corundum is found in a crystalline limestone and the detritus resulting from its decay, the limestone itself being regarded by Professor Judd as an extreme form of alteration of rocks of igneous origin (see further under Emery). Corundum has recently been reported as a constituent of both nephe- line syenites and ordinary syenites in the counties of Renfrew, Hast- ings, and Peterborough, in Eastern Ontario, Canada. According to W. G. Miller 2 these syenites are dike rocks, consisting essentially of feldspar, nepheline, and black mica or hornblende, the corundum occurring more abundantly in the ordinary syenite than in that which carries nepheline. The dikes are from a few inches to some feet in diameter, and the corundum is distributed in a somewhat capricious Bulletin No. 11. Corundum and the Basic Magnesian Rocks of Western North Carolina, by J. V. Lewis, 1896. 2 Report of the Canadian Bureau of Mims, VII, Pt. 3, 1898, p. 207. Report of U. S. National Museum, 1 899. Merir!. APPALACHIAN CRYSTALLINE .? Perldntilt-s and other Basic ;/ Mngncsian Rocks. X Corundum localities. MAP SHOWING DISTRIBUTION OF CORUNDUM AND PERIDOTITE IN THE EASTERN UNITED STATES. After.!. V. Lewis, Bulletin 11. North Carolina (Jeoloincal Snrvev. THE NONMETALLIC MINERALS. 223 manner, being quite uniformly distributed in some of the smaller dikes, or segregated irregularly along certain lines or patches. In some of the dikes the mineral is quite lacking. The total area covered by the corundum-bearing rocks, in the three counties mentioned, is 100 square miles (Specimen No. 53538, U.S.N.M.). Origin. Dr. Chatard, as a result of his observations already quoted, regards the corundum of Franklin County, North Carolina, and the Laurel Creek region of Georgia as a secondary mineral produced by a mutual reaction between the various elements of the dunite and gneiss during decomposition, the solutions formed during such decom- position giving rise to such reactions as are productive of chlorite and vermiculites, and, where the necessary conditions of proportion are reached, to corundum. On the other hand, Dr. J. H. Pratt, 1 who has made a detailed study of the North Carolina region, regards the corundum as an original constituent of the peridotite as having been held in solution in the molten magma at the time of its intrusion into the country rock, and having been one of the first minerals to crystallize on its cooling. This view is most in accord with recent synthetic work done by Moro- zewicz and others. Pirsson, who has described 2 the occurrence of sapphires in a basic eruptive rock from Yogo Gulch, Montana, regards them as of pyro- genetic origin that is, they result from the direct crystallization of the oxide, but which has been derived from aluminous material dissolved from shales by the molten rock during its intrusion. The sharp out- lines of the crystals in the granite from Gallatin County, Montana (Specimen No. 83838, U.S.N.M.), is also indicative of a direct crystalli- zation from a molten magma containing an excess of aluminum. A like origin must also be recognized for the Canadian mineral, and a part at least of that of India. EMERY. The rock emery takes its name from Cape Emeri, on the island of Naxos, where it occurs in great abundance. Mineralogically it has been regarded by various authorities as either a mechanical admixture of corundum and magnetic iron ore or as simply a massive iron spinel hercynite. So far as the Naxos emery is concerned, the first view is undoubtedly correct. Physically emery is a massive, nearly opaque, dark gray to blue-black or black material, with a specific gravity of 4 and hardness of 8, Dana's scale, breaking with a tolerably regular fracture, and always more or less magnetic. Chemically the material is quite variable in composition, a fact which gives support to the opinions of those who hold it to be a mixture rather than a true chemical compound. Below are the results of American Journal of Science, VI, 1898, pp. 49-65. 3 Idem, IV, 1897, p. 421. 224 REPORT OF NATIONAL MUSEUM, 1899. analyses by Dr. J. Lawrence Smith, from whose papers on the subject these notes are partially compiled: Localities. Alumina. Iron. Lime. Silica. Water. 61.05 27.15 1.30 9.63 2.00 I 63.50 70.10 33.25 22.21 0.92 0.62 1.61 4.00 1.90 2.10 - , 60.10 33.20 0.48 1.80 5.62 Nicaria \ 77. 82 i 71.06 8.62 20.32 1.40 4.12 2.53 I 75. 12 60.10 13.06 33.20 0.72 0.48 6.88 1.80 3.10 5.62 Ep csu 44 01 50 21 3 13 50.02 51 92 44.11 3.25 5.46 1 74.22 ( 84 02 19.31 9 63 5.48 4.81 Geologically emery, like corundum, belongs mainly to the older crystalline rocks. In Asia Minor it occurs in angular or rounded masses from the size of a pea to those of several tons weight, embedded in a blue-gray or white crystalline limestone, which overlies micaceous or hornblendic schists, gneisses, and granites. Superficial decompo- sition has, as a rule, removed more or less of the more soluble portions of the limestone, leaving the emery nodules in a red ferruginous soil. With the emery are associated other aluminous minerals as mentioned below. According to Tschermak 1 the Naxos emery (Specimen No. 60465, U.S.N.M.) occurs mostly in the form of an iron-gray, scaly to schistose, rarely massive, aggregate consisting essentially of magnetite and corun- dum, the latter mineral being in excess. In addition to these two minerals occur hematite and limonite, as alteration products of the magnetite; margarite, muscovite, biotite, tourmaline, chloritoid, dias- pore, disthene, staurolite, and rutile occur as common accessories; rarely are found spinel, vesuvianite, and pyrite. Under the microscope he finds the emery rock to show the corundum in rounded granules and sometimes well-defined crystals with hexagonal outlines, particu- larly in cases where single individuals are embedded in the iron ores. (Plate 9, fig. 2.) In many cases, as in the emery of Krenino and Pesulas, the granules are partially colored blue by a pigment some- times irregularly and sometimes zonally distributed. The corundum grains, which vary in size between 0.05 mm. and 0.52 mm. (averaging about 0.22 mm.), are very rich in inclosures of the iron ores, largely magnetite in the form of small, rounded granules. The quantity of these is so great as at times to render the mineral quite opaque, though at times of such dust-like fineness as to be translucent and of a brownish 1 Mineralogische und Petrographische Mittheilungen, XIV, 1894, p. 313. Report of U. S. National Museum, 1899. Me PLATE 9. MlCROSTRUCTURE OF EMERY. After Tscliermak. Mineralogische und Petrographische Mittheilungon, XIV. Part 4. THE NONMETALLIC MINERALS. 225 color. The larger corundums are often injected with elongated, par- allel-lying clusters or groups of the iron ores, as shown in fig. 3, Tscher- mak's paper. The corundums in turn are often surrounded by borders of very minute zircons. The iron ore, as noted above, is principally magnetite, but which, by hydration and oxidation, has given rise to abundant limonite. The magnetites are in the form of rounded granules and dust-like particles, and also at times in well-defined octa- hedrons. In their turn the magnetites also inclose particles of corun- dum very much as the metallic iron of meteorites of the pallosite group inclose the olivines and as shown in Plate 9, fig. 4. The iron ores, as a rule, occur in parallel layers and lenticular masses or nests. The following account of these deposits and the method of working is by A. Gobantz: 1 Naxos, the largest of the Cyclades Islands, is remarkable as being one of the few localities in the world producing emery on a large scale; the deposits, which are of an irregularly bedded or lenticular form, being mostly concentrated on the moun- tains at the northern end of the island, the most important ones being in the imme- diate vicinity of the village of Bothris. The island is principally made up of archsean rocks, divisible into gneiss and schist formations, the latter consisting of mica schists alternating with crystalline limestones. The lenticular masses of emery, which are very variable in size, ranging in length from a few feet to upward of 100 yards and in maximum thickness from 5 to 50 yards, are closely associated with the limestones, and, as they follow their undulations, they vary very much in position, lying at all kinds of slope, from horizontal to nearly vertical. Seventeen different deposits have been discovered and worked at different times. These range over considerable heights from 180 to 700 meters above sea-level, the largest working, that of Malia, being one of the lowest. This important deposit covers an area of more than 30,000 square metres, extending for about 500 metres in length with a height of more than 50 metres. This was worked during the Turkish occupation, and it has supplied fully one-half of all the emery exported since the formation of the Greek Kingdom. The highest quality of mineral is obtained from two comparatively thin but extensive deposits at Aspalanthropo and Kakoryakos, which are 435 metres above the sea level. The mineral is stratified in thin bands from 1 to 2 feet in thickness, crossed by two other systems of divisional planes so that it breaks into nearly cubical blocks in the working. The floor of the deposit is invariably crystalline limestone, and the roof a loosely crystalline dolo- mite covered by mica schist. The underlying limestones are often penetrated by dykes of tourmaline granite, which probably have some intimate connection with the origin of the emery beds above them. Mineralogically emery is a compact mixture of blue corundum and magnetic iron ore, its value as an abrasive material increasing with the proportion of the former constituent. This proportion has, however, been usually much overestimated. Seven samples collected by the author have been examined at the Technical High School in Vienna, and found to contain from 60 to 66 per cent of alumina. The average composition may be considered to be f corundum, the remainder being magnetite and silica in the proportion of about 2 to 1, with some carbonate of lime. The working of the deposits is conducted in an extremely primitive fashion. ^esterreichische Zeitschrift fur Berg- und Hiittenwesen, XLII, p. 143. Abstract in the Minutes and Proceedings of the Institute of Civil Engineers, CXVII, pp. 466-468. > NAT MUS 99 15 226 EEPOBT OF NATIONAL MUSEUM, 1899. During the period of Turkish rule the exclusive right of emery mining was given to two villages, and this rule has prevailed up to the present time; no Greek Govern- ment having ventured to break down the monopoly. These privileged workmen are about 600 in number, and have the right of working the mineral wherever and in what manner they may think best. The produce is taken over by the Govern- ment official at the rate of about 3 12s. for 50 cwte. The rock is exclusively broken by fire-setting. A piece of ground, about 5 feet broad and the same height, is cleared from loose material, and a pile of brushwood heaped against it and lighted. This burns out in about twenty-four or thirty hours, when water is thrown upon the heated rock to chill it and develop fractures along the secondary divisional planes in the mass of emery, and so facilitate the breaking up and removal of the material. Sometimes a crack is opened out by inserting a dynamite cartridge, but the regular use of explosives is impossible, owing to the hardness of the mineral which can not be bored with steel tools. Only the larger lumps are carried down to the shipping place, the smaller sizes, up to pieces as large as the fist, being left on the ground. As most of the suitable places for fire-setting at the surface have been worked out, attempts have been made to follow the deposits underground, but none of these have been carried to any depth, partly on account of the suffocating smoke of the fires, rendering continuous work difficult; but more particularly from the dangerous character of the loose dolomite roof, which is responsible for many fatal accidents from falls annually. These might, of course, be prevented by the judicious use of timber or masonry to support the roof, but this appears to be beyond the skill of the native miners. The rapid exhaustion of the forests in the neighbourhood of the mines, owing to the heavy consumption of fuel in fire-setting, has been a cause of anxiety to the Government for some years past, and competent experts have been employed to suggest new methods of working. These have been tolerably unanimous in recom- mending the institution of systematic quarry workings, using diamond boring machines and powerful explosives for winning the mineral, and the construction of wire-rope ways and jetties for improving the methods of conveyance and shipping; but as funds for these improvements, owing to the disastrous condition of the national finances, are not obtainable, the primitive method of working still con- tinues. Meanwhile the competition of the mines in Asia Minor has become so intense that the export of emery from Naxos has almost entirely ceased for a year past. According to Jackson, the principal emery deposit at Chester, Mas- sachusetts, in the United States, occurs at South Mountain, in the form of a bed from 4 to 10 feet in width, with a nearly N. 20 E., S. 20 W., course, and dipping to the eastward at an angle of 70. The bed widens rapidly as it rises in the mountain, and is in one place, where it is associated with a bed of iron ore (magnetite), IT feet wide, the emery itself being not less than 10 feet in the clear. The highest point of outcrop is 750 feet above the immediate base of the mountain. The bed cuts through both the South and North Mountains, and has been traced in length 4 miles. Frequently large globular masses of the emery are found in a state of great purity, separated from the principal masses of the bed and surrounded by a thin layer of bright green chloritoid and a thicker layer of interwoven laminated crystals of delicate lilac-colored margarite (Specimen No. 63107, U.S.N.M.), sometimes 2 or more inches in thickness. Some of these balls of emery are 3 or more feet in diameter and extremely difficult to break. THE NONMETALLIC MINERALS. 227 (Specimens Nos. 63102, 63103, 63104, 63105, 63106, U.S.N.M.), showthe character of the ore as mined and the character of the wall or country rock. The chief commercial sources of emery are those of Gumuch-dagh, between Ephesus and the ancient Tralles; Kulah, and near the river Hermes in Asia Minor, and the island of Naxos, whence it is quarried and shipped from Smyrna, in part as ballast, to all parts of the world. The only commercial source of importance in the United States, or indeed, in North America, is Chester, Massachusetts, as above noted. The island of Naxos is stated to have for several centuries furnished almost exclusively the emery used in the arts, the material being chiefly obtained from loose masses in the soil. The mining at Kulah and Gumuch-dagh was begun about 1847 and at Nicaria in 1850. The emery vein at Chester, Massachusetts, was discovered by Dr. H. S. Lucas in 1863, and described by Dr. C. T. Jackson in 1864. In preparing for use the mineral, after being dug from the soil or blasted from the parent ledge, is pulverized and bolted in various grades, from the finest flour to a coarse sand (Specimens Nos. 59844 to 59864, U.S.N.M., inclusive). The commercial prices vary according to grade from 3 to 10 cents a pound. At the end of the last century the price of the Eastern emery is given at from $40 to $50 a ton. About 1835 an English monopoly controlled the right of mining and the price rose in 1847 to as high as $140 a ton. The chief uses of emery and corundum, as is well known, are in the form of powder by plate-glass manufacturers, lapidaries, and stone workers; as emery paper, or in the form of solid disks made from the crushed and bolted mineral and cement, known commercially as emery wheels. The great toughness and superior cutting power of these wheels renders them of service in grinding glass, metals, and other hard substances, where the natural stone is quite inefficient. (See further under Grind and Whet Stones, p. 463.) BIBLIOGRAPHY OF CORUNDUM AND EMERY. JOHN DICKSON. Notes. American Journal of Science, III, 1821, pp. 4, 229. J. LAWRENCE SMITH. Memoir on Emery First part On the Geology and Miner- alogy of Emery, from observations made in Asia Minor. American Journal of Science, X, 1850, p. 354. J. LAWRENCE SMITH. Memoir on Emery Second part On the Minerals associated with Emery. American Journal of Science, XI, 1851, p. 53. WILLIAM P. BLAKE. Corundum in Crystallized Limestone at Vernon, Sussex County, New Jersey. American Journal of Science, XIII, 1852, p. 116. CHARLES T. JACKSON. Discovery of Emery in Chester, Hampden County, Massa- chusetts. Proceedings of the Boston Society of Natural History, X, 1864, p. 84. American Journal of Science, XXXIX, 1865, p. 87. 228 EEPOET OF NATIONAL MUSEUM, 1899. CHARLES U. SHEPARD. A Description of the Emery Mine of Chester, Hampden County, Massachusetts. Pamphlet, 16 pp., London, 1865. J. LAWRENCE SMITH. On the Emery Mine of Chester, Hampden County, Massachusetts. American Journal of Science, XLII, 1866, pp. 83-93. Original Researches in Mineralogy and Chemistry, 1884, p. 111. C. W. JENKS. Corundum of North Carolina. American Journal of Science, III, 1872, p. 301. CHARLES U. SHEPARD. On the Corundum Region of North Carolina and Georgia. American Journal of Science, IV, 1872, pp. 109 and 175. FREDERICK A. GENTH. Corundum, its Alterations arid Associated Minerals. Proceedings of the American Philosophical Society, XIII, 1873, p. 361. C. W. JENKS. Note on the occurrence of Sapphires and Rubies in situ with Corun- dum, at the Culsagee Mine, Macon County, North Carolina. Quarterly Journal of the Geological Society, XXX, 1874, p. 303. W. C. KERR. Corundum of North Carolina. Geological Survey of North Carolina, I, Appendix C, 1875, p. 64. C. D. SMITH. Corundum and its Associate Rocks. Geological Survey of North Carolina, I, Appendix D, 1875, p. 91-97. R. W. RAYMOND. The Jenks Corundum Mine, Macon County, North Carolina. Transactions of the American Institute of Mining Engineers, VII, 1878, p. 83. J. WILCOX. Corundum in North Carolina. Proceedings, Academy of Natural Sciences, Philadelphia, XXX, 1878, p. 223. F. A. GENTH. The so-called Emery-ore from Chelsea, Bethel Township, Delaware County, Pennsylvania. Proceedings, Academy of Natural Sciences, Philadelphia, XXXII, 1880, p. 311. C. D. SMITH. Corundum. Geological Survey of North Carolina, II, 1881, p. 42. F. A. GENTH. Contributions to Mineralogy. Proceedings of the American Philosophical Society, XX, 1882. A. A. JULIEN. The Dunyte Beds of North Carolina. Proceedings of the Boston Society Natural History, XXII, 1882, p. 141. T. M. CHATARD. Corundum and Emery. Mineral Resources of the United States, 1883-84, p. 714. T. M. CHATARD. The Gneiss-Dunyte Contacts of Corundum Hill, North Carolina, in Relation to the Origin of Corundum. Bulletin No. 42, U. S. Geological Survey, 1887, p. 45. G. H. WILLIAMS. Norites of the "Cortlandt Series." American Journal of Science, XXXIII, 1887, p. 194. F. A. GENTH. Contributions to Mineralogy. American Journal of Science, XXXIX, 1890, p. 47. Emery Mines in Greece. Engineering and Mining Journal, L, 1890, p. 273. A. GOBAUTZ. The Emery Deposits of Naxos. Engineering and Mining Journal, LVIII, 1894, p. 294. FRANCIS P. KING. Corundum Deposits of Georgia. Bulletin No. 2, Geological Survey of Georgia, 1894, 133 pp. T. D. PARET. Emery and Other Abrasives. Journal of the Franklin Institute, CXXXVII, 1894, pp. 353, 421. J. C. TRAUTWINE. Corundum with Diaspore, Culsagee Mine, North Carolina. Journal of the Franklin Institute, XCIV, p. 7. J. VOLNEY LEWIS. Corundum of the Appalachian Crystalline Belt. Transactions of the American Institute of Mining Engineers, XXV, 1895, THE NONMETALLIC MINERALS. J. VOLNEY LEWIS. Valuable Discovery of Corundum. Canadian Mining Review, XV, 1896, p. 230. The Corundum Lands of Ontario. Canadian Mining Review, XVII, 1898, p. 192. Corundum in Ontario. Engineering and Mining Journal, LXVI, 1898, p. 303. A. M. STONE. Corundum Mining in North Carolina. Engineering and Mining Journal, LXV, 1898, p. 490. 229 PISOLITIC BAUXITE. Bartow County, Georgia. Specimen No. 63335, U.S.N.M. 3. BAUXITE. Composition A1 2 O 3 .2H 2 O,= alumina, 73.9 per cent; water, 26.1 per cent. Commonly impure through the presence of iron oxides, silica, lime, and magnesia. Color, white or gray when pure, but yel- lowish, brown, or red through impurities. Specific gravity, 2.55; structure, massive, or earthy and clay like. According to Hayes l the 1 The Geological Relations of the Southern Appalachian Bauxite Deposits. Trans- actions of the American Institute of Mining Engineers, XXIV, 1894, pp. 250-251. 230 REPORT OF NATIONAL MUSEUM, 1899. bauxites of the Southern United States show considerable variety in physical appearance, though generally having a pronounced pisolitic structure. (See Specimens Nos. 63335, 66576, 6657T, and 66578, U.S.N.M. , from Floyd and Bartow counties, Georgia; also fig. 5, p. 229.) The individual pisolites vary in size from a fraction of a millimeter to 3 or 4 centi- meters in diameter, although most commonly the diameter is from 3 to 5 millimeters. The matrix in which they are embedded is generally more compact and also lighter in color. The larger pisolites are composed of numerous concentric shells, separated by less compact substance or even open cavities, and their interior portions readily crumble to a soft powder. In thin sections the ore is seen to be made up of amorphous flocculent grains, and the various structures which it exhibits are produced by the arrangement and degree of compactness of these grains. The matrix in which the pisolites are imbedded may be composed of this flocculent material segregated in an irregularly globular form or in compact oolites, with sharply-defined outlines. Or both forms may be present, the compact oolites being embedded in a matrix composed of the less defi- nite bodies. In some cases the interstices between the oolites are filled either wholly or in part with silica, apparently a secondary deposition. The pisolites also show considerable diversity in structure. In some cases they are composed of exactly the same flocculent grains as the surrounding matrix, from which they are separated by a thin shell of slightly denser material. This sometimes shows a number of sharply-defined concentric rings, and is then dis- tinctly separated from the matrix and the interior portion of the pisolite. The latter is also sometimes composed of imperfectly defined globular masses, and in other cases of compact, uniform, and but slightly granular substance. It is always filled with cracks, which are regularly radial and concentric, in proportion as the interior substance has a uniform texture. Branching from the larger cracks, which, as a rule, are partially filled with quartz, very minute cracks penetrate the intervening por- tions. Thus the pisolites appear to have lost a portion of their substance, so that it no longer fills the space within the outer shell, but has shrunk and formed the radial cracks. No analyses have been made of the different portions of the pisolites or of the pisolites and matrix separately, and it is impossible to say whether any differ- ences in chemical composition exist. It may be that some soluble constituent has been removed from the interior of the pisolites, but it is more probable that the shrinking observed is due wholly to desiccation. Scattered throughout the ground-mass are occasional fragments of pisolites, whose irregular outlines have been covered to varying depths by a deposit of the same material as forms the concentric shells, and thus have been restored to spherical or oval forms. Composition. The following tables will serve to show the wide range of composition of bauxites from various sources: Composition of bauxites from various localities. SiO 2 . T10 2 . A1 2 3 . FesOs. (ign) H 2 0. as? P*0 6 . Analyst. Baux, France: 1. Compact variety ' 2 8 2. Pisiform . 4 8 3 2 55 4 3. Hard and compact calcareous 30 3 34 9 paste. 4. Calabres, France 5. Thoronet, France, red variety 0.30 3.40 69.30 22.90 14 10 THE NONMETALLIC MINERALS. Composition of bauxites from various localities Continued. 231 Si0 2 . TiO 2 . A1 2 O 3 . Fe *o, gg> (100) H 2 0. P Z 6 . Analyst. 6. Villeveyrac, Herault, France, white variety 2.20 6 29 4.00 76.90 64.24 50.85 49.02 50.92 39.44 45.94 47.52 41.38 41.00 48.92 52.21 57.25 56.88 52. 13 39.75 56.10 58.61 59.82 45. 21 61.25 55.59 57.62 62.05 46.40 58.60 55.64 51.90 .10 2.40 14,36 12.90 15.70 2.27 11.86 19.95 .85 25.25 2.14 13.50 3.21 1.49 1.12 1.62 10.64 2.63 2.16 0.52 1.82 6. OS 1.83 1.60 22. 15 9.11 1.95 3.16 15 25 27.03 25.88 27.75 12.80 21.20 23 23 20.43 23.41 27 80 74 1.35 .93 .85 9.20 1.40 57 72 .65 .45 72 .46 .48 .38 trace trace .07 .... Lill. Lang. Do. Liebreich. Dr. Wm. B. Phillips. Do. Do. Do. W. F. Hille- brand. Do. Nichols. Do. Do. Prof. H. 0. White. Do. Do. Do. Do. Do. Do. Langsdorf, Germany: 5 14 9 Light red 10.27 1.10 37.87 18.67 7.73 23.72 10.25 21.08 3.20 2.53 2.52 3.52 3.60 3.55 2.08 10. Vogelsberg, Germany 11. Cherokee County, Alabama 12. Jacksonville, Calhoun County, Alabama. 13. Red 14 White 15. Red 16. White 2.80 2.30 19.56 41.47 2.56 18. Do 19. Do Georgia: 20. No. 1 21 No 2 24 10 30 31 17 31 28 28 30.31 26 28 27.62 24.86 14 10 42 10 13 43 99 63 68 63 22. No. 3 23 No 4 8.29 6.62 35.88 3.15 24. No. 5 26. No. 6 26. Barnsley estate, Dinwood Sta- tion, Georgia, No. 7. Pulaski County, Arkansas: 27. Black 28. Do 1.98 10.13 11.48 2.38 29. Do 30. Red 2.00 4.89 3.50 31 Do 3 34 32. Do 33 Do 10.38 16.76 3.50 3.50 No. 1. Contains also 0.4 CaCO 3 . No. 2. 0.2 CaCO 3 . No. 3. 12.7 CaCO 3 . No. 5. 22.90 FeO + Fe,O 3 . No. 60.10 FeO + Fe^A,. No. 7. 0.85 CaO, 0.38 MgO, 0.20 SO 3 . No. 8. 0.35 FeO, 0.41 CaO, 0.11 MgO, 0.09 K 2 O, 0.17 NaO, trace CO 2 . No. 9. FeO not det., 0.62 CaO, trace MgO, 0.11 KoO, 0.20 NaaO, 0.26 CO 2 . No. 10. 0.80 CaO, 0.16 MgO. Origin and mode of occurrence. The mineral received its name from the village of Baux, in southern France, where a highly ferrifer- ous, pisolitic variety was first found and described by Berthier in 1821. The origin of the mineral, both here and elsewhere, has been a matter of considerable discussion. The following notes relative to the foreign occurrences are from a paper by R. L. Packard: 1 The geological occurrence of the bauxite of Baux was studied by H. Coquand [Bulletin de la Society Geologique de France, XXVIII, 1871, p. 98], who describes 1 Mineral Resources of the United States, 1891, p. 148. 232 EEPOET OF NATIONAL MUSEUM, 1899. the mineral as of three varieties, pisolitic, compact, and earthy. The pisolitic variety does not differ in structure from the iron ores of Tranche Comte" and Berry, although the color and composition are different. It occurs in highly tilted beds alternating with limestones, sandstones, and clays, belonging to the upper cre- taceous period, and in pockets or cavities in the limestone. The limestone con- taining the bauxite and that adjacent thereto is also pisolitic, some nodules being as large as the fist, and the pisolitic bauxite has sometimes a calcareous cement, and at others is included in a paste of the compact mineral. M. Coquand supposed that the alumina and iron oxide composing the bauxite were brought to the ancient lake bed in which the lacustrine limestone was formed by mineral springs, which, dis- charging in the bottom of the lake, allowed the alumina and iron oxide to be dis- tributed with the other sediment. In some cases the discharge occurred on land, and the deposit then formed isolated patches. He refers to other similar deposits of bauxite of the same period in France. Sometimes the highly ferriferous mineral predominates over the aluminous (white), at others diaspose is found enveloping the red mineral, while in other cases it is mixed with it, predominating largely, and sometimes manganese peroxide replaces ferric oxide. In some places the ground was strewed with fragments of tuberous menilite, very light and white. M. Ang6 [Bull. Soc. Geolog. de France, XVI, 1888, p. 345] describes the bauxite of Var and HeVault and gives analyses of it. Over 20,000 tons were being mined in this region annually at the time of writing his report [1888]. In the red mineral of Var druses occur with white bauxite running as high as 85 per cent. Al-jOj, and 15 per cent. H 2 O, corresponding to the formula A1 2 O S +H 2 O. He refers to the prevail- ing theory of the formation of bauxite, according to which solutions of the chlorides of aluminum and iron in contact with carbonate of lime undergo double decomposi- tion, forming alumina, iron oxide, and calcium chloride. Other deposits in the south of France, in Ireland, Austria, and Italy, he says, confirm this view, because they also rest upon or are associated with limestone. The bauxite deposit in Puy de Dome which he studied could not, however, be explained by this theory because it was not associated with limestone, but rested directly upon gneiss and was partly covered by basalt. The geological sketch map of the deposit near Madriat, Puy de Dome, which he gives shows gneiss, basalt, with uncovered bauxite largely predomi- nating, and patches of miocene clay, while a geological section of the deposit near Villeveyrac, Herault, shows the bed of bauxite conformably following the flexures of the limestone formation when covered by more recent beds, and when exposed and denuded occupying cavities and pockets in the limestone. This occurrence is substantially the same as that of the neighboring Baux. M. Ang6 agrees with M. Coquand in attributing the bauxite to geyserian origin. He uses as an illustration of the contemporaneous formation of bauxite the deposits from the geysers of the Yellowstone Park, which is evidently due to a misunderstanding. He made no petrographical examination of the bauxite of Puy de Dome, nor did he attempt to trace any genetic relation between the latter and the accompanying basalt, The occurrence is, however, noteworthy, and an examination might show that it is another instance of the direct derivation of bauxite from basalt, which is maintained in the two following instances, somewhat imperfectly in the first to be sure, but with greater detail in the second. The first is a paper by Lang [in the Berichte der Deutschen Chemischen Gesell- schaft, XVII, 1884, p. 2892]. He describes the bauxite in Ober-Hessen, which is found in the fields in round masses up to the size of a man's head, embedded in a clay which is colored with iron oxide. The composition varies very widely. The petrographical examination showed silica, iron oxide, magnetite, and augite. The chemical composition and petrographical examination shows the bauxite to be a decomposition product of basalt. By the weathering of the plagioclase feldspars, augite, and olivine, nearly all the silica had been removed, together with the greater THE KONMETALLIC MINERALS. . 233 part of the lime and magnesia; the iron had been oxidized and hydrate of alumina formed as shown by its easy solubility in hydrochloric acid. The residue of the silica had crystallized as quartz in the pores of the mineral. The more detailed account of the derivation of bauxite from basalt is given in an inaugural dissertation by A. Liebreich, abstracted in the Chemisches Centralblatt, 1892, p. >4. This writer says that the well-known localities of bauxite in Germany are the bouthern slope of the Westerwald near Miihlbach, Hadamar, in the neigh- borhood of Lesser Steinheim, near Hanau, and especially the western slope of the Vogelsberg. Chemical analyses show certain differences in the composition of bauxite from different places, the smaller amount of water in the French bauxite referring it to diaspore, while the Vogelsberg mineral is probably Gibbsite (hydrar- gillite) . The bauxites of Ireland, of the Westerwald, and the Vogelsberg, show by certain external indications their derivation from basalt. The bauxite of the Vogels- berg occurs in scattered lumps or small masses, partly on the surface and partly imbedded in a grayish white to reddish brown clay, which contains also similar masses of basaltic iron ore and fragments of more or less weathered basalt itself. Although the latter was associated intimately with the bauxite, a direct and close connection of the two could not be found, but an examination of thin sections of the Vogelsberg bauxite showed that most specimens still possessed a basaltic (anamesite) structure, which enabled the author to determine the former constituents with more or less certainty. The clays from different points in the district carrying basalt, basaltic iron ore, and bauxite were examined, some of which showed clearly a sedi- mentary character. Some of the bauxite nodules were a foot and a half in diameter and possessed no characteristic form. They were of an uneven surface, light to dark brown, white, yellowish, and gray in color, speckled and pitted, sometimes finely porous and full of small colorless or yellowish crystals of hydrargillite. The thin sections showed distinct medium-granular anamesitic structure. Lath-shaped por- tions filled with a yellowish substance preponderated (the former plagioclases) and filling the spaces between these were cloudy, yellow, brown, and black transparent masses which had evidently taken the place of the former augite. Laths and plates of titanic iron, often fractured, were commonly present and the contours of altered olivine could be clearly made out. The anamesitic basalt of the neighborhood showed a structure fully corresponding with the bauxite. Olivine and titanic iron oxide were found in the clay by washing. The basaltic iron ore also showed the anamesite structure. But two localities in the United States have thus far yielded bauxite in commercial quantities. These are in Arkansas and the Coosa Valley of Georgia and Alabama. According to Branner the Arkansas beds occur near the railway in the vicinity of Little Hock, Pulaski County, and near Benton, Saline County. "The exposures vary in size from an acre to 20 acres or more, and aggregate something over a square mile." This does not, in all probability, include the total area covered by bauxite in the counties mentioned, for the method of occurrence of the deposits leads to the supposition that there are others as yet undiscovered by the survey. In thickness the beds vary from a few feet to over 40 feet, with the total thickness undetermined; the average thickness is at least 15 feet. These Arkansas deposits occur only in Tertiary areas and in the neighborhood of eruptive syenites ("granites 71 ) to which they seem 234 REPORT OF NATIONAL MUSEUM, 1899. to be genetically related. In elevation they occur only at and below 300 feet above tide level, and most of them lie between 260 and 270 feet above tide. They have soft Tertiary beds both above and below them at a few places, and must, therefore, be of Tertiary age. As a rule, however, they have no covering, the overlying beds having been removed by erosion, and are high enough above the drainage of the country to be readily quarried. Erosive action has removed a part of the bauxite in some cases, but there are, in all probability, many places at which it has not yet been even uncovered. It is pisolitic in structure, and, like all bauxite, varies more or less in color and in chemical composition. (Specimen No. 67600 from Pulaski County.) At a few places it is so charged with iron that attempts have been made to mine it for iron ore. Some of the samples from these pits assay over 50 per cent of metallic iron. This ferruginous kind is exceptional, however. From the dark red varieties it grades through the browns and yellows to pearl gray, cream colored, and milky white, the pinks, browns, and grays being the more abundant. Some of the white varieties have the chemical composition of kaolin, while the red, brown, and gray have but little silica and iron, and a high percentage of alumina. The analyses given on page 231 show that this bauxite compares favorably with that of France, Austria, and Ireland, and is apparently well adapted for the manufacture of chemical products, for refractory material, and for the manufacture of aluminum by the Deville process. The Georgia and Alabama deposits have been the subject of exhaust- ive study by Willard Hayes, to whose paper reference has already been made. According to this authority the ore is found irregularly distributed within a narrow belt of country extending from Adairsville, Georgia, southwestward, a distance of 60 miles, to the vicinity of Jackson- ville, Alabama. The only points at which it has been worked on a commercial scale are at Hermitage furnace, 5 miles north of Rome, Georgia, near Six Mile Station, south of Rome, and in the dike dis- trict near Rock Run, Alabama. (See fig. 6.) The oldest rocks of the region are of Cambrian age and are subdivided on lithologic grounds into two formations, the Rome sandstone below and the Connasauga shale above. The former consists of TOO to 1,000 feet of thin-bedded purple, yellow, and white sandstones and sandy shales. In the south- ern portion of the region the Rome sandstone is replaced by the Weisner quartzite, which consists of a series of interbedded lenticular masses of conglomerate, quartzite, and sandy shale. It apparently represents delta deposits contemporaneous with a part or the whole of the Rome sandstone. These rocks form Weisner and Indian mountains, and in the latter they attain a thickness of 10,000 feet or more. THE NONMETALLIC MINERALS. 235 The Connasauga is between 2,000 and 3,000 feet in thickness. It consists at the base of fine aluminous shales; the upper portion is more calcareous, and locally passes into heavy beds of blue seamy limestone. Above Connasauga shale is the Knox dolomite, the most uniform and persistent formation of the southern Appalachian region. It con- sists of from 3,000 to 4,000 feet of gray, semicrystalline, siliceous dolomite. The silica is usually segregated in nodules and beds of MAP SHOWING THE GEOLOGICAL RELATIONS OF THE GEORGIA AND ALABAMA BAUXITE DEPOSITS. C.W.HAYES Fig. 6. MAP SHOWING THE GEOLOGICAL RELATIONS OF GEORGIA AND ALABAMA BAUXITE DEPOSITS. After C. W. Hayes. chert. These remain upon the surface, and with the other insoluble constituents form a heavy residual mantle covering all the outcrops of the formation. It is associated with these residual materials that the extensive deposits of limonite and bauxite are found. The geo- logical structure of the region is complicated and for its details the present reader is referred to Dr. Hayes's original paper. Subaerial decomposition has progressed for a long period, and the surface is deeply covered with a mantle of residual material, consisting of the more insoluble portions of the original rock masses. This 236 EEPOBT OF NATIONAL MUSEUM, 1899. residual material consists mainly of ferruginous clay with large amounts of chert, and reaches a thickness of 100 feet or more. The bauxite deposits in the Rock Run district are regarded as typical for the entire region, and are described as follows: Four bodies of the ore were being worked in 1893 on a considerable scale, and all show practically the same form. The southernmost of the four, called the Taylor bank, is located 3 miles northeast of Rock Run, near the western base of Indian Mountain. Although the heavy mantle of residual material effectually conceals the underlying rocks, the ore appears to be exactly upon the faulted contact between the narrow belt of Knox dolomite on the northwest and the sandy shales and quartzites of Indian Mountain on the southeast. The ore is covered by 3 or 4 feet of red sandy clay in which numerous fragments of quartzite are imbedded. The ore- body is an irregularly oval mass, about 40 by 80 feet in size. Its contact with the surrounding residual clay, wherever it could be observed, appeared to be sharp and distinct, and, about the greater portion of its circumference, very nearly vertical. A certain amount of bedding is observable in the ore-body, although no trace of bedding can be detected in the surrounding residual material. Upon the northwestern or down-hill side of the ore-body, this bedding is very distinct. Layers of differently colored and differ- ently textured ore alternate in regular beds, a few inches in thickness, and above these are thinner beds of chocolate and red material, probably containing consider- ' ' DRAINAGE DITCH ' v - ~- Fig. 7. SECTION SHOWING RELATION OP BAUXITE TO MANTLE OF RESIDUAL CLAY IN GEORGIA. After C. W. Hayes. able kaolin. These beds have a steep dip, somewhat greater than the slope of the hill-side, but in the same direction. They are not simply inclined planes, however, but are curved, so as to form a steeply-pitching trough. With increasing distance from the ore-body, the lamination becomes less distinct, and the beds pass gradually into a homogeneous mottled clay. The accompanying section, fig. 7, shows these relations of the ore and residual mantle. At the Dike bank [see Fig. 6], about a mile northeast of the one above described, the stratification is well shown in portions of the deposit. Beds of yellow and gray, fiiTe-grained material, alternate with others of pisolitic ore. The beds dip at an angle of about 40, and are curved so as to form a steep trough. The compact material also shows distinct cross-bedding; both primary and secondary planes dipping in the same direction. In the Gain's Hill bank, about 250 yards north of the Dike bank, the ore-body shows a more regularly oval form than in most of the other deposits, and is also somewhat dome-shaped, swelling out laterally from the surface downward, as far as the working has progressed. Although some of the workings have gone to a considerable depth (in a few cases feet or more), the bottom of the ore-body has not been reached in any case. ore vanes in composition with depth, but not in a uniform manner, nor more to different portions at the same depth. The deepest pits have not gone e base of the surrounding residual mantle, so that no observations have yet THE NONMETALLIC MINERALS. 237 been made with regard to the relations between the ore and the country-rock; and nothing has yet been observed which warrants the conclusion that the ore if fol- lowed to sufficient depth, will be found inter-bedded with the underlying forma- tions, or even that it will be found occupying cavities in the limestone although the latter is quite possible. Concerning the origin of these deposits the author says: No eruptive rocks, either ancient or modern, are found in the vicinity of the latter, nor are there any rocks in this region which, by weathering, could yield bauxite as a residual product. Hence, any satisfactory explanation of the origin of these deposits must give the source from which the material was derived, the means by which it was transported, and the process of its local accumulation. As already stated in describing the stratigraphy of the region, the ore is associated with the Knox dolomite or with calcareous sandy shales immediately overlying the dolomite. The Connasauga, consisting of 2,000 feet or more of aluminous shales, invariably underlies the dolomite at greater or less distance beneath the ore-bearing regions, and is probably the source from which the alumina was derived. The faults of the region have been briefly described. Undoubtedly such enormous dislocations of the strata generated a large amount of heat. The fractures facilitated the circulation of water, and for considerable periods the region was probably the seat of many thermal springs. These heated waters appear to have been the agent by which the bauxite was brought to the surface in some soluble form and there precipitated. The chemical reactions by which the precipitation was effected are not well under- stood, and the conditions were not such as can be readily reproduced in the labora- tory. Of the few soluble compounds of aluminum which occur in nature, only the sulphate and the double sulphate of potash and alumina need be considered. The oxygen contained in the meteoric waters percolating at great depths through the fractured strata would readily oxidize the sulphides disseminated in the aluminous shales. Sulphates would thus be formed by a process strictly analogous to that com- monly employed in the manufacture of alum. Probably the mo.t abundant product of the process in nature was ferrous sulphate. Some sulphate of aluminum must also have been formed together with the double sulphate of potassium and aluminum, especially in the absence of sufficient potash to form alum with the whole. In its passage from the underlying shales through several thousand feet of dolo- mite the heated water must have become highly charged with lime, in addition to the ferrous and aluminous salts already in solution. But calcium carbonate reacts upon aluminum sulphate and to some extent also on alum, forming a gelatinous or floc- culent precipitate which consists of aluminum hydroxide and the basic sulphate. This reaction may have taken place at great depth and the resulting flocculent pre- cipitate may have been brought to the surface in suspension. From analogy with pisolitic sinter and travertine now forming, such conditions would appear to be highly favorable for the production of the structures actually found in the bauxite. The precipitate was apparently collected in globular masses by the motion of the ascending water, and constant changes in position permitted these to be coated with successive layers of more compact material. Finally, after having received many such coatings, the pisolites were deposited on the borders of the basin, and the interstices were filled by minute oolites formed in a similar manner or by the floc- culent precipitate itself. Slight differences in the conditions prevailing in the sev- eral springs, such as concentration and relative proportion of the various salts in solution, also temperature and flow of the water, would produce the variation in the character of the ore observed at different points. The bedding observed in the bauxite-deposits may have been produced by the successive layers deposited on the steeply inclined outlet of the basin. After the 238 REPORT OF NATIONAL MUSEUM, 1899. cessation of the spring-action, surface-creep of the residual mantle from the higher por- tions of the ridges covered the deposits to varying depths, as they are found at present. A small portion of the ferrous sulphate was oxidized and precipitated along with the bauxite, but the greater part was carried some distance from the springs and slowly oxidized, forming the widespread deposits of limonite in this region. ffggg. The better known use of bauxite is as an ore of aluminum, for which purpose it lies beyond the scope of the present work. It may, however, be well to state that before the aluminum can be satis- factorily extracted the ore is purified by chemical processes. The principal use is for the manufacture of alums and other aluminum salts such as are used in the manufacture of baking powders and dyes. It is believed that the mineral, owing to its highly refractive qualities, will in the near future be utilized in the manufacture of fire brick and crucibles. An alumino-ferric cake, a by-product obtained in the puri- fying process, is claimed as of value for sanitary and deodorizing purposes. The price of the crude ore varies greatly, according to purity. The average price for the past few years has been about $5 a ton. BIBLIOGKAPHY OF CRYOLITE AND BAUXITE. PAUL QUALE. Account of the Cryolite of Greenland. Annual Report of the Smithsonian Institution, 1866, p. 398. M. H. COQUAND. Sur les Bauxites de la chaine des Alpines (Bouches-du-Rhone) et leur age geologique. Bulletin de la Socie"te Geologique de France, 2d ser., XXVIII, 1870-71, pp. 98-115. EDWARD NICHOLS. An Aluminum Ore. Transactions of the American Institute of Mining Engineers, XVI, 1887, p. 905. P. JOHNSTRUP. Sur le Gisement de la Kryolithe au Greenland. Bulletin de la Soci ad Geological Survey of Pennsylvania, CCC, Lancaster County 1880 pp. 176 ; 192. THE NONMETALLIC MINERALS. 251 sand chrome to be found within the Baltimore region, Isaac Tyson, jr., began to fear that the sources of supply could not much longer be restricted to his ownership. In such an event he realized that he would be compelled to manufacture his ores or to sacrifice them in competition. The method of manufacture previously in use was to heat a mixture of chrome ore and potassium nitrate upon the working hearth of a reverberatory furnace. The potash salt yielded oxygen to the chromic oxide present, forming chromic acid, which, in turn, united with the base, producing potash chromate. The process was wasteful and exceedingly costly. Afterwards the process was somewhat cheapened by substitution of potassium carbonate for the more costly nitrate; oxygen was taken from heated air in the furnace. But not until 1845, when Stromeyer introduced his process, was the manufacture of chromic acid placed upon a safe mercantile basis. In this process pulverized chromic iron is mixed with potassium carbonate and freshly slaked lime, and the mixture is heated in a reverberatory furnace. After chromic oxide is set free in the charge it is freely oxidized because of the spongy conditions of the lime-laden charge. Among the first steps of Isaac Tyson, jr., was to apply, in 1846, to Yale College for a chemist for his chrome works. In response a young man named W. P. Blake, who was then a student in the chemical laboratory, was sent. For a while Mr. Blake did excellent service in the new factory, but he was not willing to remain. Mr. (now Professor) Blake was the first chemist to be employed in technology upon this continent, while the Baltimore works were the first to appreciate the value of chemistry. After the departure of Mr. Blake another chemist was secured from the first laboratory ever instituted for the teaching of chemistry, that founded at Giessen by Liebig. In succession cauie another chemist from the same laboratory, and this gentleman is yet employed in the works. Between 1880 and 1890 the American production of chrome ore has varied between 1,500 and 3,000 tons. The total eastern product in 1886 was 100 tons only. Chrome ore was discovered in California in 1873, and since 1886 this State has been the only one to produce this mineral. From 2,000 to 4,000 tons of Turkish chrome ore are now annually imported into the United States, most of which is utilized in Baltimore. BIBLIOGKAPHY. . Lake Chrome and Mineral Company, of Baltimore County. American Mineral Gazette and Geological Magazine, I. April 1, 1864, p. 253. HARRIE WOOD. Chromite and Manganese. Chromic iron and manganese ores have been found in considerable quantities, but the deposits have not yet been exten- sively worked. The chromite occurs in the Bowling Alley Point, Grafton, Young, and Bingera districts. Manganese ores are found widely distributed throughout the Colony; but the principal deposits are at Bendemere, near Moonbi, Glanmire, Rocky, and Broken Hill. Mineral Products of New South Wales, Department of Mines, 1887, p. 42. Ueber schwedisches Chromroheisen und Martinchromstahl. Berg-und Hiittenmannische Zeitung, XL VII, 1888. p. 267. Die Chromersenerz-Lagerstatten Xeuseeland. Berg-und Hiittenmannische Zeitung, XL VII, 1888. p. 375. Chrome Iron. Eighth Annual Report of the State Mineralogist of California, 1888, p. 326. Chromite Mined at Cedar Mountain. Eighth Annual Report of the State Mineralogist of California, 1888, p. 32. 252 REPORT OF NATIONAL MUSEUM, 1899. Chrome Iron Ore from Orsova. Journal of the Iron and Steel Institute, 1889, p. 316. Chrome Iron, Shasta County. Tenth Annual Report of the State Mineralogist of California, 1890, p. 638. Chromium in San Luis Obispo County. Tenth Annual Report of the State Mineralogist of California, 1890, p. 582. Chrome Iron in New Zealand. Engineering and Mining Journal, LIV, 1892, p. 393. Chromic Iron. Twelfth Report of the State Mineralogist of California, 1894, p. 35. J. T. DONALD. Chromic Iron in Quebec, Canada. Engineering and Mining Journal, LVIII, 1894, p. 224. Chromic Iron: Its Properties, Mode of Occurrence and Uses. Journal of the General Mining Association of the Province of Quebec, 1894-95, p. 108. W. F. WILKINSON. Chrome Iron Ore Mining in Asia Minor. Engineering and Mining Journal, LX, 1895, p. 4. WM. GLENN. Chrome in the Southern Appalachian Region. Transactions of the American Institute of Mining Engineers, XXV, 1895, p. 481. Chromic Iron. Thirteenth Report of the State Mineralogist of California, 1896, p. 48. GEORGE W. MAYNARD. The Chromite Deposits on Port au Port Bay, New Foundland. Transactionsof theAmericanlnstituteof Mining Engineers, XXVII, 1897, p. 283. J. H. PRATT. Chromite in North Carolina. Engineering and Mining Journal, LXVII, 1899, p. 261. The Occurrence, Origin, and Chemical Composition of Chromite, with especial reference to the North Carolina Deposits. Transactions of the American Institute of Mining Engineeers, XXIX, 1899, p. 17. 10. MANGANESE OXIDES. The element manganese exists in nature under many different forms, of which those in combination as oxides, carbonates, and silicates alone need concern us in this work. The principal known oxides are manganosite (MnO); Hausmanite (MnO,Mn 2 O 3 ); Braunite (3Mn 2 O 3 , MnSiO 3 ); Polianite (MnO 2 ) ; Pjrolusite (MnO 2 ); Manganite (Mn 2 O 3 ,H 2 O); Psilomelane (H 4 MnO 5 ); and Wad, the last being, perhaps, an earthy impure form of psilomelane. To this list should be added the mineral f ranklinite, a manganiferous oxide of iron and zinc. Of these the first named, manganosite, is rare, having thus far been reported only in small quantities associated with other oxides in Wermland, Sweden. The other forms are described somewhat in detail as below. It should be stated, however, that with the exception of the well- crystallized forms it is often diflacult to discriminate between them, as they occur admixed in all proportions, and, moreover, one variety, as pyrolusite, may result from the alteration of another (manganite). The better defined species may be separated from one another by their comparative hardness, streak, and hydrous or anhydrous properties, as shown in the accompanying table. Report of U S. National Museum, 1899. Merril PLATE 10. IDEAL SECTIONS SHOWING THE FORMATION OF MANGANESE-BEARING CLAY FROM THE DECAY OF THE ST.CLAIR LIMESTONE. MANGANE SE-BEARING CLAY LZulZARD LIMESTONE ED SACCHAROIDAL SANDSTONE sdBooNE CHERT I - i -'-'\ ST.CLAIR LIMESTONE: FIG.I. ORIGINAL CONDITION OF THE ROCKS. 1,1 FIG. 2. FIRST STAGE OF DECOMPOSITION. FIG.3. SECOND STAGET OF DECOMPOSITION. FIG. 4. THIRD STA8E OF DECOMPOSITION. SECTION SHOWING THE FORMATION OF MANGANESE DEPOSITS FROM DECAY OF LIMESTONE. After Penrose, Animal Report Geological Survey of Arkansas, I, 1K90. THE NONMETALLIC MINERALS. 253 Variety. Hardness. Specific gravity. Color. Streak. Anhydrous or hydrous. Franklinite ... 5.5 to 6.5 5 to 5.22 Iron black Reddish brown to Anhydrous. black. Hausmannite . 5 5.5 4. 7 4. 85 Brown black Chestnut brown Do. Braunite 6 6.5 4.7 4.85 Brown black to steel Brown black Do. gray. Polianite 6 6.5 4.8 4.9 Light steel gray Black Do. Pyrolusite 2 2.5 4.8 Iron black to steel Black or blue black . . a Do. gray or bluish. Manganite 4 4.2 4.4 Dark steel gray to Red brown to black . . Hydrous. iron black. Psilomelane . . . 5.6 3.7 4.7 Iron black to steel Brown black Do. gray. a Usually yields water in closed tube. The chemical relationship of the ores as found in nature is thus set forth by Penrose: 1 Chemical composition. Anhydrous form. Hydrous form. Protoxide (MnO) Manganosite (MnO) Pyrochroite (MnO.HoO). Sesquioxide (Mn 2 O 3 ) Peroxide (MnO 2 ) Braunite (Mn 2 O 3 ) Pyrolusite, Polianite ( MnO) Manganite (Mn 2 O 3 ,H 2 O). {Psilomelane. Wad. Manganese oxides frequently occur admixed in indefinite propor- tions with the hydrous oxides of iron limonite, giving rise to the manganiferous limonites as shown in Specimens Nos. 66090, 10867, U.S.N.M. from Spain. FRANKLINITE. This may be termed rather as a manganiferous ore of iron and zinc than a true ore of manganese. Nevertheless, as the residue after the extraction of the zinc is used in the manufacture of spiegeleisen, we may briefly refer to it here. The mineral occurs in rounded granules or octahedral crystals of a metallic luster and iron black color, associated with zinc oxides and silicates in crystalline lime- stones, at Franklin Furnace, New Jersey. (Specimen No. 83941, U.S.N.M.) It bears a general resemblance to the mineral magnetite, but is less readily attracted by the magnet and gives a strong manga- nese reaction. Its average content of manganese oxides Mn 2 O 3 and MnO is but from 15 to 20 per cent. HAUSMANNITE. This form of the ore when crystallized usually takes the form of the octahedron, and may be readily mistaken for franklin- ite, from which, however, it differs in its inferior hardness, lower specific gravity, and in being unacted upon by the magnet. (Specimen No. 64241, U.S.N.M.) It occurs in porphyry, associated with other Annual Report of the Geological Survey of Arkansas, I, 1890, p. 541. 254 BEPOET OF NATIONAL MUSEUM, 1899. manganese ores, in Thuringia; is also found in the Harz Mountains; Wermland, Sweden, and various other European localities. In the United States it is reported as occurring only in Iron County, Missouri. The mineral in its ideal purity consists of sesquioxide and protoxide of manganese in the proportion of 69 parts of the former to 31 of the latter. Analyses of the commercial article as mined are not at hand. BRAUNITE. This, like hausmannite, crystallizes in the form of the octahedron, but is a trifle harder. Chemically it differs, in that analyses show almost invariably from 7 to 10 per cent of silica, though as to whether or no this is to be considered an essential con- stituent it is as yet difficult to say. Analyses 1 and 2, on p. 256, show the composition of the mineral as found. The ore is reported as occurring both crystallized and massive in veins traversing porphyry at Oehrenstock in Ilmenau, in Thuringia, near Ilefeld in the Harz; Schneeberg, Saxony (Specimen No. 68136, U.S.N.M.), and various other European localities. Also at Vizianagram in India; in New South Wales, Australia, and in the Batesville region, Arkansas. POLIANITE. Like pyrolusite, yet to be noted, this form of the ore is chemically a pure manganese binoxide, carrying some 63.1 per cent metallic manganese combined with 36.9 per cent oxygen. From pyrolusite it is distinguished by its anhydrous character and increased hardness. So far as reported, it is a rather rare form of manganese, though possibly much that has been set down as pyrolusite may be in reality polianite. PYROLUSITE occurs in the form of iron black to steel gray, sometimes bluish opaque masses, granular, or commonly in divergent columnar aggregates sufficiently soft to soil the fingers, and in this respect easily separated from the other common forms excepting wad. Not known in crystals except as pseudomorphs after manganite. Its composition is quite variable, usually containing traces of iron, silica, and lime and sometimes barium and the alkalies. Analyses III and IV, on p. 256, as given by Penrose, will serve to show the general average. This is a com- mon ore of manganese, and is extensively mined in Thuringia, Mora- via, Bohemia, Westphalia, Transylvania, Australia, Japan (Specimen No. 61936, U.S.N.M.), India, New Brunswick (Specimen No. 36825, U.S.N.M.), Nova Scotia, and various parts of the United States (Specimens Nos. 42011, Tennessee, 56354, Georgia, etc.). MANGANITE differs and is readily distinguishable from the other ores thus far described, in carrying from 3 to 10 per cent of combined water, which can readily be detected when the powdered mineral is heated in a closed tube. From either psilomelane or pyrolusite it is distinguished by its hardness. When in crystals it takes prismatic forms with the prism faces deeply striated longitudinally (Specimen No. 67922, U.S.N.M., from Thuringia). Its occurrence is essentially Report of U. S. National Museum, 1899Merr PLATE 1 1 , BOTRYOIDAL PSILOMELANE, CfilMORA, VIRGINIA. Weight, 37J pounds. Specimen No. 66722, U.S.X.M. THE NONMETALLIC MINERALS. 255 the same as that of braunite. The composition of the commercial ore is given in the analyses on p. 256. PSILOMELAXE. This is, with the possible exception of pyrolusite, the commonest of the manganese minerals. The usual form of occur- rence is that of irregular nodular or botryoidal masses embedded in residual clays. It, is readily distinguished from manganite or wad by its hardness, and from hausmannite, braunite, or polianite by yielding an abundance of water when heated in a closed tube. The sample (Specimen No. 66722, U.S.N.M.), from the Crimora mines in Virginia, is characteristic. See Plate 11. The composition of the commercial ore is given in analyses V, VI, and VII on p. 256. WAD or BOG MANGANESE (Specimen No. 66602, U.S.N.M., from Cuba) is a soft and highly hydrated form of the ore, as a rule of little value, owing to impurities (analysis VIII). Asbolite is the name given to a variety of wad containing cobalt (see p. 187). See further Rhodonite and Rhodochrosite, pp. 280, 314. Origin. The deposits of manganese oxides which are of sufficient extent to be of commercial importance are believed to be in all cases of secondary origin; that is, to have resulted from the decomposition of preexisting manganiferous silicate constituents of the older crys- talline rocks and the subsequent deposition of the oxides in secondary strata. Indeed, in many instances the ore has undergone a natural segregation, owing to the decomposition of the parent rock and the accumulate of the manganese oxide, together with other difficult sol- uble constituents in the residual clay. Thus Penrose has shown 1 that the deposits of the Batesville (Arkansas) region result from the decay of the St. Clair limestone, the various stages of which are shown in the accompanying Plate 10. The fresh limestone, as shown by analy- sis, contains but 4.30 per cent manganese oxide (MnO), while the residual clay left through its decomposition contains 14.98 per cent of the same constituent. Occurrence. As above noted, the ore is found in secondary rocks, and as a rule in greatest quantities in the clays and residual deposits resulting from their breaking down. The usual form of the ore is that of lenticular masses or nodules distributed along the bedding planes, or heterogeneously throughout the clay. Penrose describes the Bates- ville ores as sometimes evenly distributed throughout a large body of clay, but in most places as being in pockets surrounded by day itself barren of ore. These pockets vary greatly in character, being some- times comparatively solid bodies separated by thin films of clay, and containing from 50 to 500 tons of ore; sometimes they consist of large and small masses of ore embedded together, and again at other times of small grains, disseminated throughout the clay. In the Crimora 1 Annual Report of the Geological Survey of Arkansas, I, 1890. 256 REPORT OF NATIONAL MUSEUM, 1899. (Virginia) deposits the ore (psilomelane) is found in nodular masses in a clay resulting from the decomposition of a shale which has been preserved from erosion through sharp synclinal folds. Bog manganese is described as occurring in an extensive deposit near Dawson settlement, Albert County, New Brunswick, on a branch of Weldon Creek, covering an area of about 25 acres. In the center it was found to be 26 feet deep, thinning out toward the margin of the bed. The ore is a loose, amorphous mass, which could be readily shoveled without the aid of a pick, and contained more or less iron pyrites disseminated in streaks and layers, though large portions of the deposit have merely a trace. The bed lies in a valley at the north- ern base of a hill, and its accumulation at this particular locality appears to be due to springs. These springs are still trickling down the hillside, and doubtless the process of producing bog manganese is still going on. 1 A bed of manganese ore in the government of Kutais, in the Caucasus, is described as occurring in nearly horizontal ly lying Miocene sandstones. The ore is pyrolusite and the bed stated as being 6 to 7 feet in thickness. Composition of manganese oxides. Constituents. Braunite. Pyrolusite. Psilomelane. Wad. I.. II. III. IV. V. VI. VII. VIII. MnO O FeO 3 87.47 9.62 86.95 9.85 90.15 88.98 84.99 10.48 80.27 14.10 63.46 25.42 2.55 0.21 1.75 CaO BaO SiO 2 0.34 0.48 0.18 0.51 4.35K 2 2.84 2.25 0.95 1.12 2.80 2.05 9.80 6.00 H 2 O 33.52 I. Batesville region, Arkansas. II. Elgersburg, Germany. III. Cheverie, Nova Scotia. IV. Cape Breton. V. Batesville region, Arkansas. VI. Schneeberg, Saxony. VII. Crimora, Virginia. VIII. Big Harbor, Cape Breton. Uses. According to Professor Penrose, 2 the various uses to which manganese and its compound are put, may be divided into three classes: Alloys, oxidizers, and coloring materials. Each of these classes includes the application of manganese in sundry manufactured products; or as a reagent in carrying on different metallurgical and chemical processes. The most important of these sources of con- sumption may be summarized as follows: 1 Anaual Report of the Geological Survey of Canada, VII, 1894, p. 146 M. 2 Annual Report of the Geological Survey of Arkansas, I, 1890. Allovs THE NONMETALLIC MINEEALS. 257 Spiegeleisen r.. , . _f 1 Alloys of manganese and iron. Ferrornanganese (^ /Alloys of manganese and copper, with or Manganese bronze.. ( without iron> {An alloy of manganese, aluminum, zinc, and copper, with a certain quantity of silicon. Alloys of manganese with aluminum, zinc, tin, lead, mag- nesium, etc. Manufacture of chlorine. Manufacture of bromine. As a decolorizer of glass (also for coloring glass, see coloring materials). Oxidizers Ag & dryer in varnishes and paints. LeClanche's battery. Preparation of oxygen on a small scale. Manufacture of disinfectants (manganates and permanganates). j-Calico printing and dyeing. Coloring glass, pottery, and brick. Coloring materials ..< ,.,-, l paints Wet Besides these main uses a certain amount is utilized as a flux in smelting silver ores, and, in the form of its various salts, is employed in chemical manufacture and for medicinal purposes. Pyrolusite and some forms of psilomelane are utilized in the manufacture of chlorine, and for bleaching, deodorizing, and disinfecting purposes. For this purpose the ore must be very pure and free from iron, lime carbonates, and alkalies. It is also utilized in the manufacture of bromine. In glass manufacture the manganese is used to accomplish two different results: First, to remove the green color caused by the presence of iron, and second, to impart violet, amber, and black colors. According to Mr. J. D. Weeks 1 the amount of manganese actually used for other than strictly metallurgical purposes in the United States is small. The value of a manganese ore depends somewhat upon the uses to which it is to be applied. Pyrolusite and psilomelane only are of value in the production of chlorine as above noted. These are rated, as stated by Penrose, according to their percentages of peroxide of manganese (MnO 2 ). The standard for the German ores is given at 57 per cent MnO 2 and 70 per cent for Spanish. For the manufacture of spiegeleisen the prices are based on ores containing not more than 8 per cent silica and 0.10 per cent phosphorus, and are subject to deductions as follows: For each 1 per cent silica in excess of 8 per cent, 15 cents a ton; for each 0.02 per cent phosphorus in excess of 0.10 per cent, 1 cent per 1 Mineral Resources of the United States, 1892, p. 178. NAT MUS 99 17 258 REPORT OF NATIONAL MUSEUM, 1899. unit of manganese. Settlements are based on analysis made on sam- ples dried at 212, the percentage of moisture in samples as taken being deducted from the weight. . The prices paid at Bessemer, Penn- sylvania in 1894, based on these percentages, were as below: Manganese. Prices per unit. Iron. Man- ganese. Cents. 6 6 6 6 Cents. 28 27 26 25 Otherwise expressed, the value ranges from $5 to $12 a ton, according to quality and condition of the market. It is probable that the total consumption in pottery and glass manu- facture does not exceed 500 tons a year, of which about two-thirds is used in glass making. The amount used in bromine manufacture and the other uses enumerated probably amounts to another 500 tons. The remainder is used in connection with iron and steel manufacture, chiefly in the production of steel and a pig iron containing considera- ble manganese for use in cast-iron car Avheels. In the crucible process of steel manufacture manganese is charged into the pots, either as an ore at the time of charging the pots or it is added as spiegeleisen or ferromanganese at the time of charging or during the melting, usually toward the close of the melting, so as to prevent too great a loss of manganese by oxidation. In the bessemer and open-hearth process the manganese is added as spiegeleisen or ferromanganese at or near the close of the process, just before the casting of the metal into ingots. It has been found in recent years that a chilled cast-iron car wheel containing a percentage of manganese is much tougher, stronger, and wears better than when manganese is absent. For this reason large amounts of manganif erous iron ores are used in the manufacture of Lake Superior pig iron intended for casting into chilled cast-iron car wheels. (See also The Mineral Industry, VIII, p. 419.) V. CARBONATES. 1. CALCIUM CARBONATE. CALCITE, CALC SPAR, ICELAND SPAR. These are the names given to the variety of calcium carbonate crystallizing in the rhombohedral division of the hexagonal system. The mineral occurs under a great variety of crystalline forms, which are often extremely perplexing to any but an expert mineralogist. The chief distinguishing characteris- tics of the mineral are (1) its pronounced cleavage, whereby it splits Report of U. S. National Museum, 1899Merrill. PLATE 12. Fig. i. \Basalt ^[Gravel Fits. 2. Fig. 3. XTheCaveD VIEWS SHOWING OCCURRENCE OF CALCITE IN ICELAND. After Thoroddsen. THE NOIOIETALLIC MINEBALS. 259 up into rhombohedral forms, with smooth, lustrous faces, and (2) its doubly refracting property, which is such that when looked through in the direction of either cleavage surfaces it gives a double image. (Specimen No. 53673, U.S.N.M.) It is to this property, accompanied with its transparency, that the mineral, as a crystallized compound, owes its chief value, though as a constituent of the rock limestone it is applied to a great variety of industrial purposes. When not suffi- ciently transparent for observing its doubly refracting properties the mineral is readily distinguished by its hardness (3 of Dana's Scale) and its easy solubility, with brisk effervescence, in cold -dilute acid. This last is likewise a characteristic of aragonite, from which it can be dis- tinguished by its lower specific gravity (2.65 to 2. 75) and its cleavage. Calcium carbonate, owing to its ready solubility in terrestrial waters, is one of the most common and widely disseminated of compounds. Only the form known as double spar, or Iceland spar, need here be considered. Origin and mode of occurrence. Calc spar is invariably a secondary mineral occurring as a deposit from solution in cracks, pockets, and crevices in rocks of all kinds and all ages. The variety used for optical purposes differs from the rhombohedral cleavage masses found in innumerable localities only in its transparency and freedom from flaws and impurities (Specimen No. 53673, U.S.N.M.). The chief commercial source of the mineral has for many years been Iceland, whence has arisen the term Iceland spar, so often applied. For the account of the occurrences of the mineral at this locality, as given below, we are indebted mainly to Th. Thoroddsen. 1 The quarry is described as situated on an evenly sloping mountain side at Reydar- fjorden, about 100 meters above the level of the ocean and a little east of the Helgustadir farm. (See Plate 12.) The veins of spar are in basalt and at this spot have been laid bare through the erosive action of a small stream called the "Silfurlakur," the Icelandic name of the spar being " Silf urberg. " The quarry open- ing is on the western side of this brook, and at date of writing was some 72 feet long by 36 feet wide (see fig. 1). In the bottom and sides of this opening the calc-spar is to be seen in the form of numer- ous interlocking, veins, ramifying through the basalt in every direction and of very irregular length and width, the veins pinching out or opening up very abruptly. In fig. 2 of plate is shown an area of some 40 square feet of the basaltic wall rock, illustrating this feature of the occurrence. Fig. 3 of the same plate shows the largest and most conspicuous vein, the smaller having been omitted in the sketch. The high cliff's on the north side of the quarry are poorer in calc-spar veins, the largest dipping underneath at an angle of about 40. 1 Geologiska Foreningens I, Stockholm Forhandlingar, XII, 1890, pp. 247-254. 260 REPORT OF NATIONAL MUSEUM, 1899. A comparatively small proportion of the calc-spar as found is fit for optical purposes. That on the immediate surface is, as a rule, lacking in transparency. Many of the masses, owing presumably to the development of incipient fractures along cleavage lines, show internal, iridescent, rainbow hues, such are known locally as "litsteinar" (lightstones). Others are penetrated by fine, tube-like cavities, either empty or filled with clay, and still others contain cavities, sometimes sufficiently large to be visible to the unaided eye, filled with water and a moving bubble. The most desirable material occurs in compara- tively small masses imbedded in a red-gray clay, filling the veinlike interspaces in the bottom of the pit. The nontransparent variety, always greatly in excess, occurs in cleavable masses and imperfectly developed rhombohedral, sometimes 1 to 2 feet in diameter, associated with stilbite. Calc-spar has been exported in small quantities from Iceland since the middle of the seventeenth century, though the business was not conducted with any degree of regularity before the middle of the present century, prior to that time everyone taking what he liked or could obtain, asking no one's permission. About the time Bartholin discovered the valuable optical properties of the mineral (in 1669), the royal parliament under Frederick III granted the necessary permission for its extraction. 1 It was not, however, until 1850 that systematic work was begun, when a merchant by name of T. F. Thomsen, at Seydisf jord, obtained permission of the owner of some three-fourths the property (the pastor Th. Erlendsson) to work the same. The quarried material was then transported on horseback to the North- fjord, and thence to Seydisf jord by water. In 1854 the factor H. H. Svendsen, from Eskif jord, leased the pastor's three-fourths right for 10 rigsdalers a year, and the remaining fourth, belonging to the Govern- ment, for 5 rigsdalers. Svendsen worked the mine successfully up to 1862, when one Tullinius, at Eskif jord, purchased the pastor's three- fourths and leased the Government's share for five years, paying there- for the sum of 100 rigsdalers [about $14 or $15]. This lease was renewed for four years longer at the rate of 5 rigsdalers per year and for the year 1872 at the rate of 100 rigsdalers, when the entire property passed into the hands of the Government in .consideration of the payment of 16,000 kroner [about $3,800]. From that time until 1882 the mine remained idle, when operations were once more renewed, though not on an extensive scale, owing, presumably in part, to the fact that Tullinius, the last year he rented the mine, had taken out a sufiicient quantity to meet all the needs of the market. Over 300 tons of the ordinary type of the spar is stated to have been sent to England and sold to "factory owners" (Fabrikanter) at about 30 kroner a ton, though to what use it was put is not stated. 1 Laws of Iceland, I, 1668, pp. 321, 322. THE NONMETALLIC MINERALS. 261 Aside from the locality at Helgustadir, calc-spar in quantity and quality for optical purposes is known to occur only at Djupifjordur, in West Iceland. The Reydharf jordhr localh"y was also visited by Mr. J. L. Hoskyns- Abrahall in the summer and autumn of 1889, and whose account 1 is reproduced in part below. Sudhrmula Sysla, of which Reydharf jordhr, the largest, bisects the east coast of Iceland, are cut out of an immense plateau, formed of horizontal sheets of volcanic rock, chiefly trachyte, between 3,000 and 4,000 feet high. This has been subsequently eroded into sharp, bare ridges with immense cliffs or steep slopes falling from them, parted by torrent valleys and fjords, the greater part of the district not reach- ing the present snow line. It is on one of these slopes, which slants down at an angle of forty degrees into Reydharf jordhr, that the unique quarry of Iceland spar is found. It consists of a cavity in the rock about 12 by 5 yards and some 10 feet high, originally filled almost entirely, but now only lined, with immense crystals, which are fitted so closely together as to form a compact mass, like a lump of sugar, with grains averaging 10 inches across. % The Syslurnadhur, 2 Jon Asmundarson Johnsen, had given me leave to examine the cave and take as many specimens as I liked, but the permission was not of very much use, there being about 5 feet of water nearly all over the bottom; and such specimens as 1 did get involved doing severe penance in walking barefoot over sharp crystals. The floor is covered with a thin layer of very fine chocolate-brown mud, which sticks as tenaciously to one's feet as to the crystals. I had to resort to tooth powder to get the latter clean, though the great heaps of spar which lie on the path side and in front of the mouth of the cave were all washed by the rain till they were as bright and trans- parent as ice. The water now running through the cave is incapable of forming calc-spar. It appears, like the surrounding rocks, to con- tain an excess of silicic acid, and either etches the surface of the spar wherever it comes in contact with it, or covers it with stilbite, the characteristic zeolite of the doleritic and basaltic rocks in Iceland. The rock in which the cave is formed is a dolerite, and darker in color than the surrounding phonolite, which is traversed by veins of black and green pitchstone. In the neighborhood df the spar it is disintegrated, colored slightly with green earth, and full of microscopic crystals of stilbite and calcite. The quany was worked till 1872 by Herra Tulinius, a Danish mer chant of Eskif jordhr. The trading station is an hour and a half's ride from Helgastadhir, the nearest farm to the quarry. (In Iceland all distances are measured in terms of the hour's ride, tima, and the day's 1 Mineralogical Magazine, IX, 1890, p. 179. 2 Magistrate, public notary, receiver of taxes, liquidator, auctioneer, etc. 262 REPORT OF NATIONAL MUSEUM, 1899. journey, leidh.) The Icelandic government in that year bought a quarter share of the quarry, and stopped the work, so that Tulinius Avas glad to sell them the rest. Five years ago an attempt was made to reopen it. One man was employed, and after spending about a week in the cave he succeeded in pumping out the water and extracting a fine block of clear spar, which was sold at a high price in London. Here, however, the work dropped, and in consequence Tulinius remains the proprietor of the whole of the calc spar that is available for physical work, and naturally sells it at a price that is calculated to make his very moderate stock last for a considerable time. 1 The reason of the Icelandic government is not very clear, but as the working of the quarry is, perhaps from patriotic motives, delegated to Herr Gunnarsson, an Icelandic merchant, whose nearest warehouse is at Seydhisf jordhr, a good day's ride from Eskifjordhr, it is hardly to be expected that the buried treasure will soon see the light. Perhaps, too, the specimens of the best quality have been already removed. Certainly clear pieces do not constitute the great mass of the spar, and if M. Labonne, who visited the cave in May, 1877 (the water being at that time frozen), could extract it "en assez grande abondance" 2 he did not leave much exposed for me to take two years later. M. Labonne speaks in his note of ramifications into the environing rock which have never been worked and suggests that this investigation might increase the impor- tance of the quarry. Such ramifications as I could see were on a very small scale. On the other hand, the thickness of the deposit has not yet been ascertained, but it is said that the best pieces occurred near the surface. For the most part the calcite is rendered semiopaque by innumerable cracks, generally following the gliding and cleavage planes ( i R and R), and apparently produced by the pressure of the spar itself, but sometimes following the conchoidal fracture. Remarkable examples of the latter kind are in the British Museum. CHALK. This is the name given to a white, somewhat loosely coherent variety of limestone composed of the finely comminuted shells of marine mollusks, among which microscopic forms known as foraminifera are abundant. The older text-books gave one to understand that f oraminif- eral remains constituted the main mass of the rock, but the researches of Sorby 3 showed that fully one-half the material was finely com- minuted shallow-water fortns, such as inoceramus, pecten, ostrea, sponge spicules, and echinoderms. Chalk belongs to the Cretaceous era, occurring in beds of varying thickness, alternating with shales, sands, and clays, and often including numerous nodules of a dark chalcedonic silica to which the name 1 It is sold by Thor E. Tulinius, Slotsholmsgade 16, Copenhagen K. "Comptes Rendus, CV., 1887, p. 1144. 3 Address to Geological Society of London, February, 1879. THE NONMETALLIC MINERALS. 268 flint is given. Though a common rock in many parts of Europe, it is known to American readers mainly for its occurrence in the form of high cliffs along the English coast, as near Dover. Until within a few years little true chalk was known to exist within the limits of the United States. According to Mr. R. T. Hill 1 there are, however, extensive beds, sometimes 500 feet in thickness, extending throughout the entire length of Texas, from the Red River to the Rio Grande, and northward into New Mexico, Kansas," and Arkansas. These chalks in many instances so closely simulate the English product, both in phys- ical properties and chemical composition, as to be adaptable to the same economic purposes. The following analyses from the report above alluded to serve to show the comparative composition: Constituents. Lower Cretace- ous chalk, Burnet County, Texas. Upper Cretace- ous chalk, Rocky Comfort, Arkan- sas. White Cliff chalk, Little River, Arkan- sas. White chalk of Shore- ham, Sus- sex, Eng- land. Gray chalk, Folk- stone, England. 92 42 88 48 94 18 98 40 94 09 Carbonate of magnesia Silica and insoluble silicates 1.38 1 59 Trace. 9 77 1.37 3 49 .08 1 10 .31 3 61 Ferric oxide and alumina Phosphoric acid, alumina, and loss .41 1.25 1.41 Trace. 1 29 Water .18 .55 .70 99.98 99.50 101 100 100 Chalk is used as a fertilizer, either in its crude form or burnt, in the manufacture of whiting (Specimen No. 26499, from Trego County, Kansas), in the form of hard lumps by carpenters and other mechanics, and in the manufacture of crayons (Specimen No. 62063, U.S.N.M.). Washed, chalk (Specimen No. 62085, U.S.N.M.) is used to give body to wall paper; as a whitewash for ceilings; as a thin coating on wood designed for gilding, being for this purpose mixed with glue; to vary the shades of gray in water-color paints, and as a polishing powder for metals. Concerning the importation and uses of chalk, Williams states: 2 Paris white is the name given to the white coloring substance prepared by grinding cliffstone, a variety of chalk or limestone which is as hard as some building stones and has a greater specific gravity than the ordinary chalk. It is imported from Hull, Eng- land, and sells at from $2 to $4 per ton ex vessel, according to freight rates from Hull. During the calendar year 1884 3,905 tons of cliffstone were imported at New York. The paris white made in this country is sold at from $1.10 to $1.25 per hundred- weight, in casks, according to make and quality. The paris white made in England, of which 508,185 pounds were imported at New York during the calendar year 1884, 1 Annual Report of the Arkansas Geological Survey, II, 1888. 2 Mineral Resources of the United States, 1883-84, p. 930. 254 REPORT OF NATIONAL MUSEUM, 1899. sells at from $1.25 to $1.30 per hundredweight. There in apparently no difference in quality between the cliffstone ground in this country and the imported paris white. Its principal use is in the preparation of kalsomine. It is also employed in the manufacture of rubber, oilcloth, wall papers, and fancy glazed papers. * Until recently all of the whiting used in this country was ground from chalk imported from Hull, P^ngland. [See Specimen No. 36013, U.S.N.M.] The annual production of whiting is about 300,000 barrels. The price varies, according to the quality, from 35 to 90 cents per hundredweight. There are four grades made, as follows: Common whiting, worth from 35 to 40 cents; gilders' whiting, 60 to 65 cents; extra gilders' whiting, 70 to 75 cents; American paris white, 80 to 85 cents. The uses of whiting are about the same as paris white, which it closely resembles. The material, as should be stated, is brought mainly as ballast from England and France. LIMESTONES; MORTARS; AND CEMENTS. Pure limestone or calcium carbonate is a compound of calcium oxide and carbonic acid in the proportion of 56 parts of lime (CaO) to 44 parts of the acid (CO 2 ). In its crystalline form as exemplified in some of our white marbles the rock is therefore but an aggregate of imperfectly outlined calcite crvstals, or, otherwise expressed, is a crystalline granular aggregate of calcite. In this "form the rock is white or colorless, sufficiently soft to be cut with a knife, and dissolves with brisk effervescence when treated with dilute hydrochloric or nitric acid. Sulphuric acid will not dissolve it except in small proportions, since the exteriors of the granules become converted shortly into insoluble calcium sulphate (gypsum), which protects them from further attack. As a constituent of the earth's crust, however, absolutely pure lime- stone is practically unknown, all being contaminated with more or less foreign material, either in the form of chemically combined or mechan- ically admixed impurities. Of the chemically combined impurities the most common is magnesia (MgO), which replaces the lime (CaO) in all proportions up to 21.7 per cent, when the rock becomes a dolomite. This in its pure state can readily be distinguished from limestone by its greater hardness and in its not effervescing when treated with cold dilute acid. (See p. 274.) It dissolves with effervescence in hot acids, as does limestone. As above noted, all stages of replacement exist, the name magnesian or dolomitic limestone being applied to those in which the magnesia exists in smaller proportions than that above given (21.7 per cent). Iron in the form of protoxide (FeO) may also replace a part of the lime. Of the mechanically admixed impurities silica in the form of quartz sand or various more or less decomposed silicate minerals, clayey and carbonaceous matter, together with iron oxides, are the more abundant. These exist in all proportions, giving rise to what are known as siliceous, aluminous or clayey, carbonaceous, and ferruginous limestones. Phosphatic material may exist in vary- ing proportions, forming gradations from phosphatic limestones to true phosphates. Limestones are sedimentary rocks formed mainly through the depo- THE NONMETALLIC MINEEALS. 265 sition of calcareous sediments on sea bottoms; many beds, however, as the oolitic limestones, show unmistakable evidences of true chemical precipitation. They are in all cases eminently stratified rocks, though the evidences of stratification may not be evident in the small speci- men exhibited in museum collections. Varietal names other than those mentioned above are given and which are dependent upon struc- tural features or other peculiarities. A shaly limestone is one partak- ing of the nature of shale. Chalk is a fine pulverulent limestone composed of shells in a finely comminuted condition and very many minute foraminifera. (See p. 262.) The name chalky limestone is fre- quently given to an earthy limestone resembling chalk. Marl is an impure earthy form, often containing many shells, hence called shell marl. An oolitic limestone 1 is one made up of small rounded pellets like the roe of a fish. The name marble is given to any calcareous or even serpentinous rock possessing sufficient beauty to be utilized for ornamental purposes. Uses. Aside from their uses as building materials, lithographic purposes, etc., as described elsewhere, limestones are utilized for a considerable variet}" of purposes, the more important being that of the manufacture of mortars and cements. Their adaptability to this purpose is due to the fact that when heated to a temperature of 1,000 F. they gradually lose the carbonic acid, becoming converted into anhydrous calcium oxide (CaO), or quicklime, as it is popularly called; and further, that this quicklime when brought in contact with water and atmospheric air greedily combines with, first, the water, forming hydrous calcium oxide (CaOH 2 O), and on drying once more with the carbonic acid of the air, forming a more or less hyd rated calcium car- bonate. In the process of combining with water the burnt lime (CaO) gives off a large amount of heat, swells to nearly twice its former bulk, and falls away to a loose, white powder. This when mixed with siliceous sand forms the common mortar of the bricklayers, or, if with sand and hair, the plaster for the interior walls of houses. (Specimens Nos. 63144, 63145, U.S.N.M.,fromVermont;No. 53195,U.S.N.M., from Maine, and No. 53168, from Pennsylvania, show the character of the rocks commonly used for these purposes.) Quicklime formed from fairly pure calcium carbonate sets or hardens after but a few days' exposure, the induration, it is stated, being due in part to crystallization. The less pure forms of limestone, notably those which contain upwards of 10 per cent of aluminous silicates (clayey matter), furnish, when burned, a quicklime which slakes much more slowly so slowly, in fact, that it is not infrequently necessary to crush to powder after burning. These same quicklimes when slaked are further differentiated from those already described by their property of setting (as the process of induration is called) under water. Hence they are known as hydraulic limes, and the rocks from which they are made as hydraulic limestones. 266 REPOKT OF NATIONAL MUSEUM, 1899. Their property of induration out of contact with the air is assumed to be due to the formation of calcium and aluminum silicates. Inas- much as these silicates are practically insoluble in water, -it follows that quite aside from their greater strength and tenacity they are also more durable; indeed, there seems no practical limit to the endurance of a good hydraulic cement, its hardness increasing almost constantly in connection with its antiquity. Certain stones contain the desired admixture of lime and clayey matter in just the right proportion for making hydraulic cement. In the majority of cases, however, it has been found that a higher grade, stronger and more enduring material, can be made by mixing in definite proportions, determined by experi- ment, the necessary constituents obtained, it may be, from widely sep- arated localities. The exact relationship Existing between composition and adaptability to lime making does not seem as yet to be fully worked out. As is well known, the pure white crystalline varieties yield a quicklime inferior to the softer blue-gray, less metamorphosed varie- ties. Nevertheless there are certain distinctive qualities, due to the presence and character of impurities, which led Gen. Q. A. Gillmore to adopt the following classification: ( 1 ) The common or fat limes, containing, as a rule, less than 10 per cent of impurities. (2) The poor or meager limes, containing free silica (sand) and other impurities in amounts varying between 10 per cent and 25 per cent. (3) The hydraulic limes, which contain from 30 to 35 per cent of various impurities. (4) The hydraulic cements, which may contain as much as 60 per cent of impurities of various kinds. As above noted most cements are manufactured from a variety of materials, and their consideration belongs therefore more properly to technology. Nevertheless it has been thought worth the while here to give in brief the matter below relative to a few of the more impor- tant and well-known varieties now manufactured. PORTLAND CEMENT. This takes its name from a resemblance of the hardened material to the well-known oolitic limestone of the island of Portland in the English Channel. As originally made on the banks of the Thames and Medway it consists of admixtures of chalk and clay dredged from the river bottoms, in the proportions of three volumes of the former to one of the latter, though these proportions may vary according to the purity of the chalk. These materials are mixed with water, compressed into cakes, dried and calcined, after which it is ground to a fine powder and is ready for use. The following analyses from Heath's Manual of Lime and Cement will serve to show the vary- ing composition of the chalk and clay from the English deposits. THE NONMETALLIC MINERALS. 267 Constituents. Upper chalk. Gray chalk. Clay. 97 90 to 98 60 87 35 to 96 52 Silica do .66 1.59 1.67 6.84 55 to 70 Magnesium carbonate do .10 .21 .10 .50 35 74 38 46 3 15 Alumina do 1.14 .93 42 4 29 11 24 3 4 Lime do 4 8 1 2 4 5 It is stated that the presence of more than very small quantities of sand, iron oxides, or vegetable matter in the clay is detrimental. A good cement mud before burning may contain from 68 to 78 per cent of calcium carbonate, 21 to 15 per cent of silica, and from 10 to 7 per cent of alumina. The following analyses from the same source as the above serve to show (I) the composition of the clay; (II) the mixed clay and chalk or "slurry," as it is called, and (III) the cement powder prepared from the same: Constituents. I. Clay. II. Slurry. III. Cement. Lime 62 13 Calcium sulphate 2.13 2 01 09 97 Silica (soluble) 54.14 11.77 20. 45 14 68 4 45 8 05 Magnesium carbonate Magnesia 4.48 2.87 1.48 Iron oxide Sand 7.76 .87 2.13 1.24 4.37 .98 Water 15 03 7 59 Several brands of Portland cement are manufactured in America, usually from a mixture of materials, the proportions of which have been worked out by experiment. At the Coplay Cement Works, in Lehigh County, Pennsylvania, a blue-gray crystalline limestone and dark gray more siliceous variety are ground and mixed into the desired proportions, molded into a brick, and burnt to the condition of a slag. The material is then ground to a powder and forms the cement. Through the courtesy of the manager, the Museum collections contain samples of the crude and manufactured materials, as follows: Lime- stone (Specimen No. 53541, U.S.N.M.); cement rock (No. 53542, U.S.N.M.). Composition formed by admixing the two rocks (No. 53543, U.S.N.M.); and the clinker (No. 53544, U.S.N.M.) obtained by 268 KEPORT OF NATIONAL MUSEUM, 1899. burning the composition. The chemical composition of the sampl< as given are as follows: Constituents. Limestone. Cement rock. Compound of the two. Clinker. 2.10 15.22 13.22 22.74 } .84 4.24 5.20 10.50 Calcium carbonate (CaCO 3 )... Magnesian carbonate (MgCO 3 ) 96.17 Trace. 69.88 4.60 77.00 4.20 C&O61.82 MgO2.05 An impure limestone, forming a portion of the water-lime group of the Upper Silurian formations at Buffalo, New York, forms a "natural cement" rock which is utilized in the manufacture of the so-called Buffalo Portland cement. 1 The so-called Rosendale cement is- made from the tentaculite or water limestones of the Lower Helderburg group as developed in the township of Rosendale, Ulster County, New York. According to Darton 2 there are two cement beds in the Rosendale- Whiteport region, at Rosendale the lower bed or dark cement averaging some 21 feet in thickness and the upper or light cement 11 feet, with 14 to 15 feet of water-lime intervening. In the region just south of Whiteport the upper white cement beds have a thickness of 12 feet and the lower or gray cement of 18 feet, with 19 to 20 feet of water- lime beds between them. The underlying formation is quartzite. The method of mining the material from the two beds, as well as their inclination to the horizon, is shown in Plate 13. (See Specimens, Nos. 63062-63086, U.S. KM., from Ulster, Onondaga, and Erie Counties, New York; Nos. 63090-63099, U.S.N.M., Cumberland and Hancock, Maryland; No. 53173, from Lisbon, Ohio, and No. 53193, from Sandusky, Ohio). ROMAN CEMENT. The original Roman cement appears to have been made from an admixture of volcanic ash or sand (pozzuolana, pepe- rino, trass, etc.) and lime, the proportions varying almost indefinitely according to the character of the ash. The English Roman cement is made by calcining septarian nodules dredged up from the bottoms of Chichester Harbor and off the coast of Hampshire, and from similar nodules obtained from the Whitby shale beds of the Lias formations in Yorkshire and elsewhere. The following analysis of the cement stone from Sheppey, near South End, will serve to show the character of the material: dement Rock and Gypsum Deposits in Buffalo. J. Pohlman. Transactions of the American Institute of Mining Engineers, XVII, 1889, p. 250. 2 Report of the State Geologist of New York, 1, 1893. Report of U. S. National Museum, 1 899. Merrill. PLATE 13. THE NONMETALLIC MINERALS. 269 Carbonate of lime 64. 00 Silica 17. 75 Alumina 6. 75 Magnesia 50 Oxide of iron 6. 00 Oxide of manganese 1. 00 Water 3.00 Loss... 1.00 100.00 The names concrete and beton are applied to admixtures of mortar, hydraulic or otherwise, and such coarse materials as sand, gravel, fragments of shells, tiles, bricks, or stone. According to Gillmore the matrix of the beton proper is a hydraulic cement, while that of the concrete is nonhydraulic. The terms are, however, now used almost S} 7 nonymously. Aside from their uses as above indicated limestones are used in the preparation of lime for fertilizing purposes. For this purpose, as before, the lime carbonate is reduced to the condition of oxide by burn- ing, and then allowed to become air slaked, when it remains in the condition of a fine powder suitable for direct application to the land as is the plaster made from gypsum. A lime prepared by burning oyster shells is utilized in a similar manner. BIBLIOGRAPHY. Out of the many hundreds of titles that might be given, a few only are selected. Those desiring may find a very full bibliography in a series of papers on The Chemi- cal and Physical Examinations of Portland Cement. Journal of the American Chemical Society, XV and XVI. 1893-1894. Q. A. GILLMORE. Practical Treatise on Limestones, Hydraulic Cements, and Mortars. New York, 1863, 333 pp. The Cement Works on the Lehigh. Second Pennsylvania Geological Survey, Lehigh District, D. D. 1875-76, p. 59. HENRY C. E. REID. The Science and Art of the Manufacture of Portland Cement with Observations on some of its Constructive Applications. London, 1877. JOHANN BIELENBERG. Method for Utilizing Siliceous Earths and Rocks in the Manu- facture of Cements, for the purpose of imparting to them Hydraulic Properties. (German Patent No. 24038, November 28, 1882.) Journal of the Society of Chemical Industry, III, 1884, p. 110. U. CUMMINGS. Hydraulic Cements, Natural and Artificial, their Comparative Values. Massachusetts Institute of Technology, November, 1887. M. H. LE CHATELIER. Recherches Experimental sur la Constitution des Mortiers. Hydrauliques. Chas. Dunod, Paris, 1887. M. A. PROST. Note sur la Fabrication et les Proprietes des Ciments de Laitier. Annales des Mines, XVI, 1889, p. 158. H. PEARETH BRUMELL. Natural and Artificial Cements in Canada. Science, XXI, 1893, p. 177. M. H. LE CHATELIER. Precedes d'Essai des Materiaux Hydrauliques. Annales des Mines, IV, 1893, p. 367. 270 REPORT OF NATIONAL MUSEUM, 1899. A. H. HEATH. A Manual of Lime and Cement. London, 1893, 215 pp. G. R. REDGRAVE. Calcareous Cements: Their Nature and Uses. London, 1895, 222 pp. URIAH CUMMINGS. American Cements. Boston, 1898, 299, pp. CHARLES D. JAMESON. Portland Cement, its Manufacture and Use. New York, 1898, 192 pp. BERNARD L. GREEN. The Portland Cement Industry of the World. (Reprinted from Journal of the Association of Engineering Societies. XX, June, 1898). PLAYING MARBLES. At Oberstein on the Nahe, Saxony, playing marbles are made in great quantities from limestone. The stone is broken into square blocks, each of such" size as to make a sphere the size of the desired marble. These cubes are then thrown into a mill consisting of a flat, horizontally revolving stone with numerous con- centric grooves or furrows on its surface. A block of oak of the same diameter as the stone and resting on the cubes is then made to revolve over them in a current of water, the cubes being thus reduced to the spherical form. The process requires but about fifteen minutes. LITHOGRAPHIC LIMESTONE. For the purpose of lithography there is used a fine-grained homogeneous limestone, breaking with an imper- fect, shell-like or conchoidal fracture, and as a rule of a gray, drab, or yellowish color. A good stone must be sufficiently porous to absorb the greasy compound which holds the ink and soft enough to work readily under the engraver's tool, yet not too soft. It must be uniform in texture throughout and be free from all veins and inequalities of any kind, in order that the various reagents used may act upon all exposed parts alike. It is evident, therefore, that the suitability of this stone for practical purposes depends more upon its physical than chemical qualities. An actual test of the material by a practical lithographer is the only test of real value for stones of this nature. Nevertheless the analyses given below are not without interest as showing the variation in composition even in samples from the same locality. THE NONMETALLIC MINERALS. 271 li I II -5^ s K 8 | | | | g E H S f II 272 REPORT OF NATIONAL MUSEUM, 1899. Localities. Stones possessing in a greater or less degree the proper qualities for lithographic purposes have from time to time been reported in various parts of the United States; from near Bath and Stony Stratford, England; Ireland; Department of Indre, France, and also Silesia, India, and the British American possessions. By far the best stone, and indeed the only stone which has as yet been found to satisfactorily fill all the requirements of the lithographer's art, and which is the one in general use to-day wherever the art is practiced, is found at Solenhofen, near Pappenheim, on the Danube, in Bavaria. (Specimens Nos. 35888 and 35706, U.S.N.M.) These beds are of Upper Jurassic or Kimmeridgian age and form a mass some 80 feet in thickness, though naturally not all portions are equally good, or adapted for the same kind of work. The stone varies both in texture and color in different parts of the quarry, but the prevailing tints are yellowish or drab. In the United States materials partak- ing of the nature of lithographic stone have been reported from Yavapai County, Arizona (Specimens Nos. 62798 and 68162, U.S.N.M.); Talladega County, Alabama; Arkansas; Lawrence County, Indiana (Specimen No. 25030, U.S.N.M.); near Thebes and Anna, Illinois (Specimens Nos. 61344 and 62570, U.S.N.M.); James and Van Buren counties, Iowa; Hardin, Estelle, Kenton, Clinton, Rowan, Wayne, and Simpson counties, Kentucky (Specimen No. 36897, U.S.N.M., from Simpson County); near Saverton, Rails County, Missouri (Specimen No. 28498, U.S.N.M.); Clay and Overton counties, Tennessee; Burnet and San Saba counties, Texas (Specimens Nos. 38624 and 70671, U.S.N.M.); near Salt Lake City, Utah, and at Fincastle, Virginia. While, however, from nearly, if not quite every one of these localities, it was possible to get small pieces which served well for trial purposes, each and every one has failed as a constant source of supply of the commercial article, and this for reasons mainly inherent in the stone itself. It is very possible that ignorance as to proper methods of quarrying may have been a cause of failure in some cases. The Arizona stone is one of the most recent discoveries, and accord- ing to first reports seems also the most promising. Samples of the stone submitted to the writer, as well as samples of work done upon it, seemed all that could be desired (Specimens Nos. 62798 and 68162, U.S.N.M.). It is stated by Mr. W. F. Blandy that the quarries are situated on the east slope of the Verdi Range, about 2 miles south of Squaw Peak and at an elevation of about 1,200 feet above the Verdi Valley, 40 miles by wagon road east of Prescott. Two quarries have thus far been opened in the same strata, about 1,000 feet apart, the one showing two layers or beds 384 feet in thickness, and the other three beds 3,188 feet in thickness. As at present exposed the beds, which are of Carboniferous age, are broken by nearly vertical fissures into blocks rarely 4 or 5 feet in length. Owing to the massive form of THE NONMETALLIC MINERALS. 273 the beds and this conchoidal fracture the stone can not be split into thin slabs, but must be sawn. No satisfactory road yet exists for its transportation in blocks of any size, and such material as has thus far been produced is in small slabs such as can be ' ' packed. " Those who have inspected the properties express themselves as satisfied that blocks of good size and satisfactory quality can be had in quantity. The Alabama stone as examined by the writer is finely granular and too friable for satisfactory work. Qualitative tests showed it to be a siliceous magnesian limestone. It is of course possible that the single sample shown does not fairly represent the product. The Arkansas deposit is situated in Township 14 N., R. 15 W. of the 5th p. m., sections 14, 23, and 24, Searcy County. The color is darker than that of the Bavarian stone. The reports of those who have tested it are represented as being uniformly favorable. The Illinois stone is darker, but to judge from the display made in the Illinois building at the World's Columbian Exposition, 1893, is capable of doing excellent work and can be had in slabs of good size (Specimens Nos. 61344 and 62570, U.S.N.M.). The Kentucky stone is hard and brittle, though that from Rowan County is stated to have received a medal at the exposition of 1876. It is fine grained and homogeneous and very pure, only a small flocoulent residue of organic matter remaining insoluble in dilute hydrochloric acid. The Indiana stone is harder than the Bavarian, and samples exam- ined were found not infrequently traversed by fine, hard veins of calcite. (Specimen No. 25030, U.S.N.M.) The stone from Saverton, Missouri, is compact and fine grained, with, however, fine streaks of calcite running through it. (Specimen No. 28498, U.S.N.M.) It leaves only a small brownish residue when dis- solved in dilute acid. This stone has been worked quite successfully on a small scale. The State geologist, in writing on the subject, says: 1 Some of the beds of the St. Louis limestone (Subcarboniferous) have been success- fully used for lithographic work. No bed is, however, uniformly of the requisite quality, and the cost of selection of available material would seem to preclude the development of an industry for the production of lithographic stone. From the deposit at Overton, Tennessee, it is stated slabs 40 by 60 inches by 3J inches thick were obtained, though little, if anything, is now being done. An analysis of this stone is given in the table. Other promising finds are reported from McMinn County, in the same State. According to the State geological reports, the stone lies between two beds of variegated marble. The stratum is thought to run entirely through the county, but some of the stone is too hard for lithographic purposes. The best is found 8 miles east of Athens on the farm of Robert Cochrane, and a quarry has been opened by a Cincinnati company, which only pays a royalty of $250 per annum. It is sold for nearly the same price as the Bavarian stone. It is a cal- careous and argillaceous stone, formed of the finest sediment, of uniform texture, and possesses a pearl-gray tint. The best variety of this stone has a conchoidal fracture and is free from spots of all kinds. Bulletin No. 3, Geological Survey of Missouri, 1890, p. 38. NAT MUS 99 18 274 REPORT OF NATIONAL MUSEUM, 1899. A lithographic stone is described in the State survey reports of Texas as occurring at the base of the Carboniferous formations near Sulphur Springs, west of Lampasas, on the Colorado River, and to be traceable by its outcrops for a distance of several miles, the most favorable showing being near San Saba. The texture of the stone is good; but as it is filled with fine reticulating veins of calcite (Specimen No. 70671, U.S.N.M.), and as moreover the lithographic layer itself is only some 6 or 8 inches in thickness, it is obvious that little can be expected from this source. The Texas Lithographic Stone Company, with headquarters at Burnet, have used the stone, it is said, in con- siderable quantities. A stone claiming many points of excellence has for some years been known to exist in the Wasatch range within a few miles of Salt Lake City, and several companies are or have been engaged in its exploitation. Very encouraging reports of beds examined by men whose opinions should be conservative, come from Canadian sources, and it is possible a considerable industry may yet be here developed, though little is being done at present. The descriptions as given in the geological reports are as follows: l The lithographic stones of the townships of Madoc and Marmora and of the counties of Peterboro and Bruce have been examined and practically tested by lithographers, and in several cases pronounced of good quality; they have also obtained medals at various exhibitions. They were obtained from the surface in small quarries, and possibly when the quarries are more developed better stones, free from "specks" of quartz and calcite, will be available in large slabs. It should be stated that in actual use the principal demand is for stones some 22 or 28 by 40 inches; the largest ones practically used are some 40 by 60 inches and 3 to 3i inches thick. As the better grades bring as high as 22 cents a pound, it will be readily perceived that the field for exploration is one offering considerable inducement. 2. DOLOMITE. This is a carbonate of calcium and magnesium (Ca, Mg), CO 3 , = calcium carbonate 54.35 per cent, magnesium carbonate 45.65 per cent. Hardness 3.5 to 4; specific gravity, 2.8 to 2.9; colors when pure, white, but often red, green, brown, gray, or black from impurities. (Specimen No. 82167, attached crystals on limestone from Joplin, Mis- souri.) Dolomite, like calcite, occurs in massive beds or strata either compact (Specimen No. 37795, U.S.N.M.) or coarsely crystalline, and is to the eye alone often indistinguishable from that mineral. Like limestone, the dolomites occur in massive forms suitable for building purposes, or in some cases as marble. (Specimen No. 25075, U.S.N.M.) From the limestone they may be distinguished by their increased hard- ness and being insoluble in cold dilute hydrochloric acids. The dolo- mites, like the limestones, are sedimentary rocks, though it is doubtful 1 Geology of Canada, 1863. THE NONMETALLIC MINERALS. 275 if the original sediments contained sufficient magnesium carbonate to constitute a true dolomite. They are regarded rather as having resulted from the alteration of limestone strata by the replacement of a part of the calcium carbonate by carbonate of magnesium. Uses, Aside from its use as a building material, dolomite has of late come into use as a source of magnesia for the manufacture of highly refractory materials for the linings of converters in the basic processes of steel manufacture. According to a writer in the Industrial World 1 the magnesia is obtained by mixing the calcined dolomite with chloride of magnesia, whereby there is formed a soluble calcic chloride which is readily removed by solution, leaving the insoluble magnesia behind. According to another process the calcined dolomite is treated with dissolved sugar, leading to the formation of sugar of lime and deposi- tion of the magnesia; the solution of sugar of lime is then exposed to carbonic acid gas, which separates the lime as carbonate, leaving the sugar as refuse. Recently it has been proposed to use magnesia as a substitute for plaster of paris for casts, etc. The snow-white coarsely crystalline Archean dolomite commercially known as snowflake marble, and which occurs at Pleasantville, in West- chester County, New York (Specimen No. 30863, U.S.N.M.), is finely ground and used as a source of carbonic acid in the manufacture of the so-called soda and other carbonated waters. (Specimen No. 3080-1, U.S.N.M.) 3. MAGNESITE^ This is a carbonate of magnesium, MgCO 3 , = carbon dioxide 52.4 per cent, magnesia 47.6 per cent. Usually contaminated with carbon- ates of iron and free silica. The following analysis will serve to show the average run of the ma- terial, both in the crude state and after calcining: Constituents. Styria. Greece. Crude magncsite. Carbonate of magnesia 90. to 96. 94. 4G Carbonate of lime 0. 5 to 20 4 40 3 to 60 FeO 08 Silica 1 52 5 Water 54 Burnt magncgite. 77 6 82 46 to 95 36 7 3 83 to 10 92 Alumina and ferric oxide 13.0 0.56 to 3.64 Silica 1 2 73 to 7. 98 The mineral occurs rarely in the form of crystals, but is commonly in a compact finely granular condition of white or yellowish color some- Junel, 1893. 276 REPORT OF NATIONAL MUSEUM, 1899. what resembling unglazed porcelain (Specimen No. 16070, from Gilroy, California), and more rarely crystalline granular, like limestone 'or dolomite (Specimen No. 48273, U.S.N.M., from Wells Island). It is hard (3.5 to 4.5) and brittle, with a vitreous luster, and is unacted upon by cold, but dissolves with brisk effervescence in hot hydrochloric acid. Localities and mode of occurrence. Most commonly the mineral is found in the form of irregular veins in serpentinous and other magne- sian rocks, being a decomposition product either of the serpentine itself or of the original rock from which the serpentine is derived. It is also found in granular aggregates disseminated throughout serpen- tinous rocks. It is stated by Dana to occur associated with gypsum. Prof. W. P. Blake has described 1 immense beds of very pureinag- nesite as occurring in the foothills of the Sierra Nevadas, between Four and Moore creeks, in what is now Tulare County. The beds are from 1 to 6 feet in thickness and are interstratified with talcose and chloritic schists and serpentine. Mr. H. G. Hanks, who has since inspected these deposits, reports them as existing in several hills or low moun- tains, the mineral cropping out boldly in distinct and clearly marked veins, varying from 2 inches to 4 feet, and of a maximum length, as exposed, of 500 feet. In section 5, T. 15 S., R. 24 E., Fresno County, California, there is stated 2 to be a large vein of the material averaging 10 feet in width, incased in hornblendic and micaceous shales. A white marble-like crystalline granular variety has been found in the form of drift bowlders on an island in the St. Lawrence River near the Thousand Islands Park. (Specimen No. 48273, U. S. N. M. ) According to Canadian geologists magnesite forming rock masses occurs associ- ated with the dolomites, serpentines, and steatites of the eastern town- ships of Quebec. In Bolton it occurs in an enormous bed resembling crystalline limestone in appearance. An analysis of this yielded: Car- bonate of magnesia, 59.13 per cent; carbonate of iron, 8.72 per cent; silica, 32. 20 per cent. In the township of Sutton a slaty variety yielding as high as 80 per cent of carbonate of magnesium occurs admixed with feldspar and green chromiferous mica. In Styria the material lies in Silurian beds consisting of argillaceous shales, quartzites, dolomites, and limestones, resting upon gneiss. The extensive deposit of mag- nesite occurring associated with Subcarboniferous limestones in the Swiss Tyrol is regarded by M. Koch 8 as due to a decomposition of the original limestone through percolating magnesia-bearing solutions. Magnesia being the stronger base replaces the lime, which is carried away in solution. The chief localities of magnesite, native and foreign, are as fol- 1 Pacific Railroad Reports, V, p. 308 2 Tenth Annual Report of the State Mineralogist of California, 1890, p. 185. 3 Zeitschrift der Deutschen Geologischen Gesellschaft, XLV, Pt. 2, 1893, p. 294. THE NONMETALLIC MINERALS. 277 lows: Maryland, Bare Hills, Baltimore County. New Jersey, Hobo- ken. Massachusetts, Roxbury. New York, near Rye, Westchester County; Warwick, Orange County; Stony Point, Rockland County; New Rochelle, Westchester County; Serpentine Hills, Staten Island. North Carolina, Webster, Jackson County ; Hamptons, Yancey County, McMakins Mine, Cabarrus County. Pennsylvania, Goat Hill, West Nottingham, Chester County; Scotts Mine, Chester County; Low's Chrome Mine, Lancaster County (Specimen No. 53101, U.S.N.M.). California, Coyote Creek, near Madison Station, Southern Pacific Railroad, Santa Clara County (Specimen No. 16070, U.S.N.M.); Gold Run, Iowa Hill, and Damascus, Placer County; Arroyo Sero, Monterey County; Mariposa and Tuolumne counties; Diablo Range, Alaineda County; between Four Creek and Moores Creek, near Visalia, Tulare County (Specimen No. 63842, U.S.N.M.); Alameda County; Napa County (Specimen No. 62594, U.S.N.M.); Millcreek, Fresno County. Washington, Spokane County(Specimen No. 53235, U.S.N.M.). Sutton, Quebec, lot 12, range 7; Bolton, Quebec. Regla, near Havana, Cuba. Kongsberg, Norway. Piedmont, Italy. Bingera Diamond Fields, New South Wales. Victoria, South Australio (Specimens Nos. 28466 and 28472, U.S.N.M.). Kosewitz and Frankenstein, Silesia. Styria, in Austria-Hungary. Greece (Specimens Nos. 62895 and (J7983, U.S.N.M.). Uses. Magnesite is used in the preparation of magnesian salts (Epsom salts, magnesia, etc.), in the manufacture of paint, paper, and fire brick. For the last-named purpose it is said to answer admirably, particularly where a highly refractive material is needed, as in the so-called basic process of iron smelting. Magnesia made from the carbonate [magnesite] by driving off the carbonic acid is very refractory, if pure. It is made into any shape that is required, and is one of the most refractory of substances. It was formerly very difficult to get the carbon- ate of magnesia, but large quantities of it have been found on the island of Eubcea, so that it can now be had for $15 to $25 per ton, instead of $60 to $70 as formerly. It can be calcined at a lesa cost than ordinary lime, losing half of its weight, so that if calcined before it is transported the cost may be still further reduced. It contains a little lime, silicates of iron, and some serpentine and silica. After calcination, the serpentine and silica can be separated, as it is easily crushed, but the most of the work can be done by hand-picking beforehand. Before moulding, it must be sub- mitted to about the temperature it is to undergo in the furnace, otherwise it would contract. It is then mixed with a certain portion of less calcined material, which is one-sixth for steel fusion, and 10 to 15 per cent, water by weight, and pressed in iron moulds. If for any reason either because there was too much or too little water, or because the material was not properly mixed, or contains silica the crucible is not strong enough, it has only to be dipped in water, which has been saturated with boracic acid, and then heated. 1 Twenty or more years ago the mineral was mined from serpentinous J T. Egleston, Transactions of the American Institute of Mining Engineers, IV, 1876, p. 261. 278 REPOBT OF NATIONAL MUSEUM, 1899. rocks in Lancaster County, Pennsylvania, by McKim, Sines and Com- pany, of Baltimore, by whom it was used for the manufacture of Epsom salts (sulphate of magnesia). Although it is said 1 that these gentlemen made a pure salt at less price than it could be imported, and thereby excluded the foreign material almost exclusively, the mines are now wholly abandoned. Isaac Tyson & Co., of this same city, also operated mines in Lan- caster County. Early in the fall of 1886 a small force of men was set to work on the deposits of magnesite discovered on Cedar mountain, Alameda county, California. Since that time several carloads of the mineral have been gotten out and shipped by rail to New York, these deposits being only a few miles from the line of the Central Pacific Railroad. The mineral occurs here in a decomposed serpentine rock and in a yellow clay in which are embedded large bowlders. It lies in pockets and small veins, the latter running in every direction. The richest spots are found under the bowlders, where the mineral is quite pure. A machine is used to sift out the small stones from the powdered magnesite, a good deal of which is met with. A number of veins of this mineral has been exposed by the occurrence of landslides on the side of the mountain where they are situated; only a few of them, however, contain good mineral, nor is there any certainty as to how long these will last. The claims are being opened by tunnels, of which two have been started. The process of gathering this mineral is slow, as every piece has to be cleaned by hand and the whole has to be carefully assorted according to purity. Having been divided into three classes, it is put up in sacks weighing from 80 to 100 pounds each. This sacking is preliminary not only to shipping but to getting it down from the mountains, which can be done only on the backs of animals. While carbonate of magnesia occurs at a great many places in California and elsewhere on the Pacific coast, the above is the only deposit of this mineral that is being worked. An artificial article of this kind is obtained as a by-product in the manufacture of salt by the Union Pacific Salt Company of California. 2 Th. Schlossing has proposed 3 to utilize magnesian hydrate obtained by precipitation from sea water by lime, for the preparation of fire brick, the hydrate being first dehydrated by calcination at a white heat, after which it is made up into brick form. According to the Industrial World* magnesite as a substitute for barite in the manufacture of paint is likely to prove of importance. The color, weight, and opacity of powder add to its value for this purpose. In Europe it is stated the material is used as an adulterant for the cheaper grades of soap. Prices. During 1892 the material, 96 to 98 per cent pure, was quoted as worth $9 to $15 a ton in New York City. Material containing as high as 15 to 30 per cent silica and 8 to 10 per cent of iron is said to be practically worthless. In 1899 crude California magnesite was quoted as worth $3 a ton at the mines. 1 Report C. C. C. Second Geological Survey of Pennsylvania, p. 178. 2 Mineral Resources of the United States, 1886, p. 696. 3 Comptes Rendus, 1885, p. 137. industrial World, XXXVI, No. 20, 1891. THE NONMETALLIC MINEKALS. 279 4. WlTHERITE. This is a carbonate of barium of the formula BaCO 3 , = baryta 77.7 per cent, carbon dioxide 22.3 per cent. Color, white to yellow or gray, streak white; translucent. Hardness, 3 to 3.75; specific gravity, 4.29 to 4.35. When crystallized, usually in form of hexagonal prisms, with faces rough and longitudinally striated. Common in globular and botr} T oidal forms, amorphous, columnar, or granular in structure. The powdered mineral dissolves readily in hydrochloric acid, like calcite, but is easily distinguished from this mineral by its great weight and increased hardness, as well as by its vitreous luster and lack of rhomboidal cleavage, which is so pronounced a feature in calcite. From barite, the sulphate of barium, with which it might become confused on account of its high specific gravity, it is readily distinguished by its solubility in acids as above noted. From stronti- anite it can be distinguished by the green color it imparts to the blowpipe flame. Localities and mode of occurrence. The mineral occurs apparently altogether as a secondary product filling veins and clefts in older rocks and often forming a portion of the gangue material of metalliferous deposits. The principal localities as given by Dana are Alston Moor, Cumberland (Specimen No. 67923, U.S.N.M.), where it is associated with galena. In large quantities at Fallowfield near Hexain in North- umberland; at Anglezarke in Lancashire; at Arkendale in Yorkshire, and near St. Asaph in Flintshire, England. Tarnowitz, Silesia; Szlana, Hungary; Leogang in Salzburg; the mine of Arqueros near Coquimbo', Chile; L. Etang Island; near Lexington, Kentucky, and in a silver-bearing vein near Rabbit Mountain, Thunder Bay, Lake Superior. Uses. The mineral has been used to but a slight extent in the arts. As a substitute for lime it has met with a limited application in mak- ing plate glass, and is also said to have been used in the manufacture of beet sugar, but is now being superseded by magnesite. 5. STEONTIANITE. This is a carbonate of strontium, SrCO 3 , = Carbon dioxide 29.9 per cent; strontia 70.1 per cent. Often impure through the presence of carbonates and sulphates of barium and calcium. Colors, white to gray, pale green, and yellowish. Hardness 3.5 to 4. Specific gravity 3.6 to 3.7. Transparent to translucent. When ciystallized often in acute spear-shaped forms. Also in granular, fibrous, and columnar globular forms. Soluble like calcite in hydrochloric acid, with effer- vescence, but readily distinguished by its cleavage and greater density. The powdered mineral when moistened with hydrochloric acid and held on a platinum wire in the flame of a lamp imparts to the flame a very characteristic red color. 280 REPORT OF NATIONAL MUSEUM, 1899. Occurrence. According to Dana the mineral occurs at Strontian in Argyllshire, in veins traversing gneiss, along with galena and barite; in Yorkshire, England; at the Giants Causeway, Ireland; Clausthal, in the Harz; Braunsdorf, Saxony; Leogang, in Salzburg; near Brix- legg, Tyrol; near Hamm and Minister, Westphalia. In the United States, at Schoharie, New York, in the form of granular and columnar masses and also in crystals, forming nests and geodes in the hydraulic limestone; at Clinton, Oneida County; Chaumont Bay and Theresa, Jefferson County; and Mifflin County, Pennsylvania. lf seSt Strontianite, so far as the writer has information, has but a limited application in the arts. It is stated 1 that " basic bricks " are prepared from it by mixing the raw or burnt strontianite with clay or argillaceous ironstone in such proportions that the brick shall contain about 10 per cent of silica, and then working into a plastic mass with tar or some heavy hydrocarbon. After molding, the bricks are dusted with fine clay or ironstone, dried, and burned. The effect of the dusting is to form a glaze on the surface, which protects the brick from the moisture of the air. Like celestite, it is also used in the pro- duction of the red fire of fireworks. The demand for the material is small, and the price but from $2.50 to $4 a ton. 6. RHODOCHROSITE; DIALOGITE. This is a pure manganese carbonate of the formula MnCO 3 ,= carbon dioxide, 38.3 per cent; manganese protoxide, 61.7 per cent. The color is much like that of rhodonite (see p. 314), from which, however, it is readily distinguishable by its rhombohedral form, inferior hard- ness (3.5 to 4.5), and property of dissolving with effervescence in hot hydrochloric acid, while rhodonite is scarcely at all attacked. The mineral is a common constituent of the gangue of gold and silver ores, as at Butte, Montana; Austin, Nevada, etc. (Specimen No. 26T45, U.S. KM.) So far as known the mineral has as yet no commercial value. 7. NATRON, THE NITRUM OF THE ANCIENTS. This is a hydrous sodium carbonate; Na 2 CO 3 +10H 2 O, = carbon dioxide, 15.4 per cent; soda, 21.7 per cent; water, 62.9 per cent. Occurs in nature, according to Dana, only in solution, as in the soda lakes of Egypt and elsewhere, or mixed with other sodium carbonates. The artificially crystallized material is of white color when pure, soft and brittle, and with an alkaline taste. Crystals, thin, tabular, monoclinic. Thermonatrite, also a hydrous sodium carbonate of the formula Na 2 CO 3 +H 2 O= carbon dioxide, 35.5 per cent; soda, 50 percent, and water 14.5 per cent, occurs under similar conditions, and is con- sidered as derived from natron as a product of efflorescence. (See further under Sodium sulphates, p. 405.) Journal of the Society of Chemical Industry, III, 1884, p. 33. THE NONMETALLIC MINERALS. 281 8. TRONA; URAO. This is a hydrous sodium carbonate, corresponding to the formula Na 2 CO 3 .HNaCO 3 +2H 2 O,=carbon dioxide, 38.9 per cent; soda, 41.2 per cent; water, 19.9 per cent. Found in nature as an efflorescence or incrustation from the evapo- ration of lakes, particularly those of arid regions. W. P. Blake has recently described 1 crude carbonate of soda (Trona) occurring in the central portion of a basin-shaped depression or dry lake in southern Arizona, near the head of the Gulf of California. The deposit covers an area of some 60 acres to a depth of from 1 to 3 feet, the lower portion being saturated with water from a solution so strong that when exposed to the air soda is deposited at the rate of an inch in thickness for every ten days. In its native condition the soda is naturally somewhat impure, from silt blown in from the surrounding land. The analysis given below shows the general average: Sand, silt, etc 13.00 Iron oxides and alumina 2. 80 Lime 1.14 Salt(NaCl.?) 4.70 Sulphate of soda 4. 70 Carbonate of soda . . . . 73. 66 100. 00 See further under Thernardite, p. 415. VI. SILICATES. 1. FELDSPARS. The name feldspar is given to a group of minerals resembling each other in being, chemically, silicates of aluminum with varying amounts of lime and the alkalies potash and soda. All members of the group have in common two easy cleavages whereby they split with even, smooth, and shining surfaces along planes inclined to one another at angles of nearly if not quite 90. (Specimen No. 67361, U.S.N.M.) They vary from transparent through translucent to opaque, the opaque form being the more frequent. In colors they range from clear and colorless through white and all shades of gray to yellowish, pink and red, more rarely greenish. On prolonged exposures to the weather they become whitish and opaque, gradually decomposing into soluble carbonates of lime and the alkalies, and soluble silica, any one of which may be wholly or in part removed by percolating waters, leaving behind a residual product, con- sisting essentially of hydrous silicates of alumina, to which the names kaolin and clay are given (see p. 325). The hardness of the feld- spars varies from 5 to 7 of Dana's scale; specific gravity 2.5 to 2.8. 1 Engineering and Mining Journal, LXV, 1898, p. 188. 282 BEJfOBT OF NATIONAL MUSEUM, They are fusible only with difficulty, and with the exception of the mineral quartz are the hardest of the common light-colored, minerals. From quartz they are readily distinguished by their cleavage charac- teristics noted above. Geologically the feldspars belong to the gneisses and eruptive rocks of all ages, certain varieties being characteristic of certain rocks and furnishing important data for schemes of rock classi- fication. Nine principal varieties are recognized, which on crystallo- graphic grounds are divided into two groups. The first, crystallizing in the monoclinic system, including only the varieties orthoclase and hyalophane; the second, crystallizing in the triclinic system, including microclinic, anorthoclase, and the albite-anorthite series, albite, oligo- clase, andesine, labradorite, and anorthite. The above-mentioned properties are set forth in the accompanying table. Constituents. Ortho- clase. Hyalo- phane. Micro- cline. Anorth- oclase. Alhite. Oligo- clase. Ande- sine. Labra- dorite. Anor- thite. Silica SiO 2 Alumina A1 2 O 3 Potash KO 64.7 18.4 16.9 51.6 21.9 10.1 64.7 18.4 16.9 66.0 20.0 5.0 8.0 68.0 20.0 62.0 24.0 60.0 26.0 53.0 30.0 43.0 37.0 Soda Na^O 12.0 9.0 8.0 4.0 Barite BaO 16.4 LimeCaO ' 5.0 2.56-2.7 6. 0-7. 7.0 2.6-2.7 5.0-6.0 13.0 2.6-2.7 6.0 20.0 2.6-2.8 6.0-7.0 Specific gravity 2.4-2.6 6.0-6.5 2.8 6.0-6.5 2.4-2.6 6.0-6.5 2.0-5.8 2.5-2.6 6.0-7.0 Crystalline system. . . Monoclinic. Triclinic. . Of the above those which most concern us here are the potash feld- spars orthoclase and microcline, two varieties which for our purposes are esssentially identical, both as regards composition and general physical properties as well as mode of occurrence. Indeed, although crystallizing in different systems they are as a rule indistinguishable but by microscopic means or by careful crystallographic measurements. Occurrence. The feldspars are common and abundant constituents of the acid rocks such as the granites, gneisses, syenites the ortho- clase and quartzose porphyries, and the tertiary and modern lavas such as trachyte, phonolite, and the liparites. Among the older rocks they not infrequently occur in large veins or dike-like masses of coarse pegmatitic crystallization, the individual crystals being in some cases a foot or more in diameter. The asso- ciated minerals are quartz and white mica, with beryl, tourmaline, garnet, and a great variety of rarer minerals. The ordinary white mica of commerce comes from deposits of this nature and often the two minerals are mined contemporaneously. Such of our feldspars as have yet been worked for economic purposes occur associated only with the older rocks the granites and gneisses of the Archean and Lower Paleozoic formations. Near Topsham, Maine, is one of these pegmatitic veins, running THE NONMETALLIC MINEKALS. 283 parallel with the strike of the gneissoid schists in which it lies, i. e. northeast and southwest. The vein material is quartz, feldspar, and mica. The quarry, as described by R. L. Packard, is in the form of an open cut in the hillside, being some 300 feet long by 100 feet wide, and of very irregular contours. The present floor and the sides of the cut are of feldspar (Specimen No. 61086, U.S.N.M.), containing irreg- ular bodies of quartz and mica, the first named occurring in large masses entirely free from other minerals, though a second grade is taken out which is in reality an intimate mixture of quartz and feldspar. The quartz occurs, besides as mentioned above, in the form of irreg- ular bodies, sometimes 6 or 8 feet across and 15 feet or more long. It also occurs in cavities, or geodes, in the form of flattened crystals (Specimen No. 61085, U.S.N.M.). The mica is hereof little economic importance, being found in the mass of the feldspar and along the seams in the form of narrow, lanceolate masses, often arranged in small radi- ating conical forms with their apexes outward. The principal feldspar quarries thus far worked are in the Eastern United States, from Maine to New Jersey. The material is mined from open cuts, being blasted out with powder and separated from adhering quartz, mica, and other minerals by hand, after which it is shipped in the rough to the potteries,, or in some cases ground and bolted in the near vicinity. In Connecticut the material has in times past been ground by huge granite disks mounted like the wheels of a cart on an axle through the center of which extended a vertical shaft. By the slow revolution of this shaft the wheels traveled around in a limited circle over a large horizontal granite slab. The pieces of spar being placed upon the horizontal slab were thus slowly ground to powder, after which it was bolted and sacked. The modern method of pulver- izing is by means of the so-called "Cyclone" crusher. The value of the uncrushed material delivered at the potteries is but a few dollars a ton. Hence, while there are unlimited quantities of the material in different parts of the Appalachian region, but few are so situated as to be profitably worked. Uses. The feldspars are used mainly for pottery, being mixed in a finely pulverized condition with the kaolin or cjay. When subjected to a high temperature the feldspar fuses, forming a glaze and at the same time a cementing constituent. There are other substances more readily fusible which are utilized for this purpose in the cheaper kinds of ware, but it is stated that in the highest grades of porcelain, as those of Sevres, feldspar is the material used. The proportions used vary with different manufacturers, each having adopted a formula best adapted for his own workings. 2. MICAS. Under this head are comprised a number of distinct mineral species, alike in crystallizing in the nionoclinic system and having a highly 284 REPOKT OF NATIONAL MUSEUM, 1899. perfect basal cleavage, whereby they split readily into thin, trans- lucent to transparent, more or less elastic sheets. Chemically they are in most cases orthosilicates of aluminum with potassium and hydrogen, and in some varieties magnesium, ferrous, and ferric iron, sodium, lithium, and more rarely barium, manganese, titanium, and chrornium. Seven species of mica are commonly recognized, of which but three have any commercial value, though a fourth form, lepidolite, may perhaps be utilized as a source of lithia salts. Of these three forms, the white mica, muscovite, and the pearl gray, phlogopite, are of greatest importance, the black variety, biotite, being but little used. Muscovite, or potassium mica, is essentially a silicate of aluminum and potassium, with small amounts of iron, soda, magnesia, and water. Its color is white to colorless, often tinted with brown, green, and violet shades. When crystallized it takes on hexagonal or diamond-shaped forms, as do also phlogopite and biotite as shown in samples (Speci- mens Nos. 62377 and 30763, U.S.N.M.). Its industrial value lies in its great power of resistance to heat and acids, its transparency, and its wonderful fissile property, in virtue of which it may be split into extremely thin, flexible sheets. It has been stated, though I know not how correctly, that sheets but one two hundred and fifty thou- sandths (1/250000) of an inch in thickness have been obtained. Phlog- opite, or magnesian mica, differs from muscovite in being of a darker, deep pearl gray, sometimes smoky, often yellowish, brownish red, or greenish color. Biotite, or magnesia iron mica, differs in being often deep, almost coal black and opaque in thick masses, though trans- lucent and of a dark brown, yellow, green, or red color in thin folia. It further differs from the preceding in that its folia are less elastic, and the sheets of smaller size. Lepidolite, a lithia mica, is much more rare than either of the above, is of a pale rose or pink color, folia usually of small size, commonly occurring in scaly granular forms without crystal outlines. The following table will serve to show the varying composition of the four varieties mentioned: Variety. SiO 2 A1 2 3 Fetf, FeO MgO CaO K 2 Na^O F. H 2 Muscovite Phlogopite 45.71 44.48 45.40 39.66 43.00 40.64 44.94 34.67 39.30 40.16 50.39 49.62 36.57 35.70 33.66 17.00 13.27 14.11 31.69 30.09 16.95 15.79 28.19 27.30 1.19 1.09 2.36 0.27 1.71 2.28 4.75 2.42 0.48 2.53 1.07 1.07 0.20 0.69 3.90 16.99 8.45 4.12 0.71 Trace. 1.86 26.49 27.70 27.97 0.46 0.10 9.22 9.77 8.33 9.97 10.32 8.16 8.00 7.55 7.79 7.64 12.34 11.19 0.79 2.41 1.41 0.60 0.30 1.16 0.59 1.57 0.49 0.37 2.17 0.12 0.72 0.69 2.24 5.67 0.82 0.93 0.28 0.89 5.15 5.45 4.83 5.50 5.46 2.99 0.78 3.21 3.85 4.64 4.02 3.58 2.36 1.52 Biotite Lepidolite 1.98 21.89 26.15 5.08 4.34 0.82 Li 2 Li 2 0.31 0.07 THE NONMETALLIC MINEKALS. 285 Although the basal cleavage which permits of the ready splitting of the mica into thin sheets is the only one sufficiently developed to be of economic importance, the mica as found is often traversed by sharp lines of separation, called gliding planes, which may, by their abun- dance, be disastrous to the interests of the miner. Such partings, or gliding planes, supposed to be induced by pressure, are developed at angles of about 66i with the cleavage, and may cut entirely through a block or extend inward from the margin only a short distance and come to an abrupt stop. In many cases the mica is divided up into long narrow strips, from the breadth of a line to several inches in thickness, with sides parallel, and as sharply cut as though done with shears. (Specimens Nos. 62517, 63134, U.S.N.M.) The imperfections in mica are due to inclosures of foreign minerals, as flattened garnets, to the presence of free iron oxides, often with a most beautiful dendritic structure, to the partings or gliding planes noted above, and to crumplings and V-like striations which destroy its homogenity. (Specimens Nos. 63139, 44450, U.S.N.M.) Occurrence. Mica in quantity and sizes to be of economic impor- tance is found only among the older rocks of the earth's crust, par- ticularly those of the granite and gneissoid groups. Muscovite and biotite are among the commonest constituents of siliceous rocks of all kinds and ages, while phlogopite is more characteristic of calcareous rocks. It is, however, only when developed in crystals of consider- able size in pegmatitic and coarsely feldspathic veins, or, in the case of phlogopite, in gneissic and calcareous rocks associated with erup- tive pyroxenites, that it becomes available for economic purposes. The associated minerals are almost too numerous to mention. The more common for muscovite are quartz and potash feldspar, which form the chief gangue materials in crystals and crystalline masses, sometimes a foot or more in diameter. With these are almost inva- riably associated garnets, beryls, and tourmalines, with more rarely cassiterite, columbite, apatite, fluorite, topaz, spodumene, etc. In- deed, so abundant are, at times, the accessory minerals in the granitic veins, and so perfect their crystalline development, that they furnish by far the richest collecting grounds for the mineralogists. Of these minerals the quartz and feldspars are not infrequently contemporane- ously with the mica and utilized in the manufacture of pottery and abrasives. The origin of these pegmatitic veins is a matter of considerable doubt. The finer grained pegmatites are, in certain cases, undoubted intrusives, though to some authorities it seems scarcely possible that the extremely coarse aggregates of quartz, feldspar, and mica, with large garnets, beryls, and tourmalines, can be a direct result of cooling from an igneous magma. To such it seems "more probable that they are portions of an original rock mass altered by exhalations of fluor- 286 REPORT OF NATIONAL MUSEUM, 1899. hydric acids, like the Saxon "greisen." Others regard them as resulting from the very slow cooling of granitic material injected in a pasty condition, brought about by aqueo-igneous agencies, into rifts of the preexisting rocks. It must be remembered that the high degree of dynamic metamorphism which these older rocks have undergone render the problems relating to their origin extremely difficult. Localities. From what has been said regarding occurrences, it is evident that mica deposits are to be looked for only in regions occupied by the older crystalline rocks. In the United States, therefore, we need only look for them in the States bordering immediately along the Appalachian range and in the Granitic areas west of the front range of the Rocky Mountains. 1 In the Appalachian region south of Canada mica mines, worked either for mica alone or for quartz and feldspar in addition, have from time to time been opened in various parts of Maine, New Hampshire, Connecticut, Maryland, Virginia, North Carolina, and perhaps other States, but in none of them, with the exception of New Hampshire and North Carolina, has the business proven sufficiently lucrative to warrant continuous and systematic working. Indeed, were it not for the increased demand lately arising for the use of mica in electrical machines it is doubtful if any but the most favorably situated mines would remain longer in operation in the United States. This for the reason not so much that foreign mica is better as that it is cheaper. In Maine muscovite has been mined in an intermittent manner along with quartz and feldspar at the well-known mineral localities at Paris Hill and Rumford, Oxford County; Auburn, Androscoggin County; Topsham, Sagadahoc County; Edgecomb, Lincoln County, and other counties in the southeastern part of the State. In New Hampshire the industry has assumed greater importance. The mica-bearing belt is described by Prof. C. H. Hitchcock as usually about 2 miles in width, and extending from Easton, in Graf ton County, to Surry, in Cheshire County; being best developed about the towns of Rumney and He- bron. The mica occurs in immense coarse granite veins in a fibrolitic mica schist (Specimen No. 63029, U.S.N.M.) of Montalban age, and is found in sheets sometimes a yard in length, but the more common sizes are but 10 or 12 inches in length. Immense beryls, sometimes a yard in diameter, and beautiful large tourmalines occur among the accessory minerals. Mines for mica were opened at Graf ton as early as 1840, and as many as six or eight mines have been worked at one time, though by no means continually. Other mines have been worked in Groton, Alexandria, Graf ton, and Alstead, in Graf ton Country; Acworth and Springfield, Sullivan County; Marlboro, Cheshire County; New Hampton, Belknap County, and Wilmot, Merrimack County, though only those of Groton are in operation at date of writing (1894). 1 The region of the Black Hills of South Dakota is an important exception. THE NONMETALLIC MINERALS. 287 As seen by the writer, the veins at the latter place cut sharply across the fibrolitic schist, and though the vein materials adhere closely to the wall rock on either side, without either selvage or slickensides, still the line of demarcation is perfectly sharp. There seems no room for doubt but that the vein material was derived by injection from below, though from their extremely irregular and universally coarsely crystal- line condition we must infer that the condition of the injected magma was more in the nature of solution than fusion, as the word is ordinarily used, and also that the rate of cooling and consequent crystallization was very slow. The feldspars not infre- quently occur in huge crystalline masses several feet in diameter, though sometimes more finely intercrystallized with quartz in the form known as pegmatite. [Specimen No. 62519, U.S.N.M.] The mica is by no means disseminated uniformly throughout the vein, but on the contrary is very sporadic, and the process of mining consists merely in following up the mineral wherever indications as shown in the face of the quarry are sufficiently promising. Most of the mines are in the form of open cuts and trenches, though in a few instances underground cuts have been made for a distance of a hun- dred feet or more. The mica blocks as removed are of a beautifully smoky-brown color, and often show a distinct zonal structure, indicating several periods of growth. The associated feldspar is not in all cases orthoclase, but, as at the Alexandria mines, sometimes a faintly greenish triclinic variety. Samples of the New Hampshire micas, with the accompanying gangue and wall rocks, are shown in Specimens Nos. 02515 to (32519 and 63028 to 63030, U.S.N.M. In Connecticut some mica (muscovite) has been obtained in connec- tion with the work of mining feldspar and quart/ in and about the towns of Haddam, Glastonbury, and Middletown, but the business has never assumed any importance. Mica mines have also been worked in Montgomery County, Maryland. South of the glacial limit mica mining has proven more successful from the reason that the gangue minerals (feldspar and quartz) were in a state of less compact aggre- gation, due to weathering, the feldspar being often reduced to the state of kaolin, and hence readily removed by pick and shovel. The following account of the deposits of North Carolina is given by Prof. W. C. Kerr: 1 I have stated elsewhere, several years ago, that these veins were wrought on a large scale and for many ages by some ancient peoples, most probably the so-called Mound Builders; although they built no mounds here, and have left no signs of any permanent habitation. They opened and worked a great many veins down to or near water-level; that is, as far as the action of atmospheric chemistry had softened the rock so that it was workable without metal tools, of the use of which no signs are apparent. Many of the largest and most profitable of the mines of the present day are simply the ancient Mound Builders' mines reopened and pushed into the hard unde- com posed granite by powder and steel. Blocks of mica have often been found half imbedded in the face of the vein, with the tool-marks about it, showing the exact limit of the efficiency of those prehistoric mechanical appliances. As to the geolog- ical relations of these veins, they are found in the gneisses and schists of the Archaean horizons. * * * These rocks are of very extensive occurrence in North Carolina, constituting in faot the great body of the rocks throughout the whole length of the State, some 400 miles east and west, being partially covered 1 Transactions of the American Institute of Mining Engineers, VIII, 1880, p. 457. 288 REPORT OF NATIONAL MUSEUM, 1899. up, and interrupted here and there by belts of later formation. Mica veins are found here, in fact may be said to char- acterize this horizon everywhere, from its eastern outcrop, near the seaboard, to and quite under the flanks of the Smoky Mountains. It is, however, in the great plateau of the west, between the Blue Ridge and the Smoky, that the mica veins reach their greatest development, and have given rise to a very new and profitable industry, new and at the same time very old. It may be stated as a very gen- eral, almost universal, fact, that the mica vein is a bedded vein. Its position (as to strike and dip) is dependent on and controlled by, and quite nearly conformable to, that of the rocks in which it occurs, and hence, as well as on account of their great size," some observers, accustomed to the study of veins and dikes and the characters of intrusive rocks in other regions, have been disposed to question the vein character of these masses at first. But a good exposure of a sin- gle one of them is generally suffi- cient to remove all doubt on this score. The mica vein is simply and always a dike of very coarse granite. It is of any size and shape, from a few inches generally a few feet to many rods (in some cases several hundred feet) in thickness, and in length from a few rods to many hundred yards, extending in some cases to half a mile or more. The strike, like that of the inclosing rocks, is generally northeast, and the dip southeast, at a pretty high angle; but they are subject, in these respects, to many and great local variations, all the conditions being occasionally changed, or even re- versed. An idea may be formed of the coarseness of these veins from this statement, that the masses of cleavable feldspar and of quartz (limpid, pale yellow, brown, or, more generally, slightly smoky), and of mica, are often found to measure several yards in two or three of their dimen- sions, and weighing several tons. I have a feldspar crystal from one of these mines THE NONMETALLIC MINERALS. 289 of nearly a thousand pounds weight, and I have known a single block of mica to make two full two-horse wagon-loads, and sheets of mica are sometimes obtained that measure three and four feet in diameter. There are many peculiarities about these veins. Among the most important, in a practical sense, is the arrangement of the different constituents of the vein inter se, Sometimes the mica, for example, will be found hugging the hanging- wall; some- times it is found against both walls; again it may be distributed pretty equally through the whole mass of the vein; sometimes, again, it will be found collected in the middle of the vein; in other cases, where the vein varies in thickness along its course, the mica will be found in bunches in the ampullations, or bellies, of the vein; in still other cases, where the vein has many vertical embranchments, the mica will be found accumulated in nests along the upper faces of these processes or offshoots. Those features of structure will be best understood from a few repre- sentative diagrams. Figure 9 is a horizontal section, with several transverse vertical sections, of a typ- ical vein in Yancey County, at the Presnel Mine. The length of the section, i. e., of the portion of the vein that has been stripped, is 125 feet; the thickness varies from 3 to 10 feet, except at a few points, as b c where it is nearly 20 feet. The crystals of mica are found in this mine generally near the foot wall, in the recesses or pouches; at c, however, as seen in section D, it is found next the hanging- wall. The inclosing rock in this case is a hard, gray slaty to schistose gneiss. * * * The feldspar, which constitutes the larger part of the mass of these veins, is often found converted into beds of the finest kaolin; and, curiously enough, this was one of the first and most valuable exports to England in the early part of the seventeenth century, "packed" by the Indians out of the Unaka (Smoky) Mountains, and sold under the name "unakeh" (white). This kaolin, like the mica, will doubtless soon come again into demand, after lying forgotten for generations. Characteristic samples of the micas of the region are shown in Specimens Nos. 18205, 18207, 62962, and 62964, U.S.N.M. In Alabama, along a line stretching from Chilton County, north- east, through Coosa, Clay, and Cleburne counties, there are numer- ous evidences of prehistoric mica mining. Many pits are met with around which pieces of mica are still to be seen. In some places, just as in Mitchell County, North Carolina, large pine trees have grown up on the debris, so that a considerable time must have elapsed since the mines were worked. About ten years ago, Col. James George, of Clanton, Chilton Count}', prospected for mica, and some fairly good specimens were obtained, but the investigations were not continued. It is not thought that any mica has been marketed from Alabama. The indications of good mica along the line mentioned aie, however, sufficient to warrant additional and more extended examinations. Lit- tle mica is reported from other Southern States, though some mines have been opened in South Carolina, Georgia (Specimens Nos. 63139 to 63141, U.S.N.M.), Virginia, and West Virginia, In 1881, a mica mine was opened in Anderson County, South Carolina, and some miners from Mitchell County, North Carolina, employed. The enterprise was not successful, and the miners returned home shortly afterwards. Good mica has been found in South Carolina, notably in Anderson, Oconee, and Pickens counties. The mica-bearing rocks of western North Caro- NAT MUS 99 19 290 REPORT OF NATIONAL MUSEUM, 1899. lina do not protrude into Tennessee, except at intervals, and then only for short distances. Some prospecting has been done in Tennes- see near Roan Mountain, but the results were not considered satis- factory. 1 In Colorado mica has long been known to be widely disseminated and to occur in many places in bodies of workable size, but mining has until lately always proved the mica to be "plumose" and unfit for cut- ting into sheets. Many mines have been located, but the product has always proved worthless, until in the summer of 1884 the Denver Mica Company opened a mine near Turkey Creek, about 35 miles from Den- ver. This mica is of fair quality, and quite a considerable quantity of it has been mined. It is slightly brown and the largest plates which have yet been cut are not more than 2f by 6 inches in size. Only an extremely small percentage of the gross weight is available for cutting into sheets. An effort is being made to put it upon the market, and at present four workmen are employed in trimming the sheets. Mica of good quality and in large plates has also been recently reported from the neighborhood of Fort Collins. In Wyoming, mica has been found in workable quantities near Dia- mond Park and in the Wind River country, as well as at many points along the mountain ranges in Laramie County. It has recentty been mined to some extent at Whalen canon, 20 miles north of Fort Lara- mie, and some of the product has been shipped to the Eastern market. In New Mexico mica occurs near Las Vegas, and reports of ship- ments have been published. At Petaca, the Cribbenville mica mines are being worked at present by sixteen men. Work was commenced at these mines July 2, 1884, and the amount of excavation at present is 13,160 cubic feet. The plates cut range from 2 by 2 inches to 5 by 8 inches in size. Some specimen plates have been cut 10 by 12 inches, but the general average is about 3 by 4 inches. Some 12 tons of mica have been handled, but the amount sold and the average price obtained are not reported. Other localities in New Mexico also yield mica, but none have been developed, except the two above mentioned. (Specimen No. 61335, U.S.N.M.). In California many deposits of mica have been noted, especially at Gold lake, Plumas county; in Eldorado county; Ivanpah district, San Bernardino county; near Susanville, Lassen county, and at Tehachapi pass, Kern county. In 1883 a large deposit was discovered in the Salmon mountains, in the northwestern part of the State, and some prospecting was done. 2 The mica-bearing deposits of the Black Hills of South Dakota hyve been variously regarded by different observers as intrusive granites or true segregation veins lying parallel to the apparent bedding. New- 1 Mineral Resources of the United States, 1887, p. 671. 2 Idem, 1883-84, p. 911. THE NONMETALLIC MINERALS. 291 ton and Jenny, 1 Blake, 2 and Vincent regard them as intrusive, while Carpenter 3 and Crosby 4 hold the opposite view. According to Blake the mica occurs in granitic masses, remarkable for the coarseness of their crystallization, the constituent minerals being usually large and separately segregated. " Large masses of pure quartz are found in one place and masses of feldspar in another, and the mica is often accumulated together instead of being regularly dis- seminated through the mass. It also occurs in large masses or crys- tals, affording sheets broad enough for cutting into commercial sizes." Associated with the mica at this point are the minerals quartz and feldspar, mainty a lamellar albite (Clevelandite), which form the gangue, and irregularly disseminated cassiterite (tinstone), gigantic spodumenes, black tourmalines, and, in small quantities, block mica, beryls, garnets, columbite, and a variety of phosphatic minerals, such as apatite, tri- phylite, etc. In Nevada mines have been worked in the St. Thomas mining dis- trict, Lincoln County, the mica occurring in hard, glassy quartz rock forming an outcrop some 200 feet wide by 600 feet long in gneiss and schists. At the Czarina Mine, located in May, 1891. near Rioville, the mica occurs under similar conditions. The mineral seems to follow the division plane of the stratification, along the line or axis of fold. This line runs north and south, slightly east of north of the main trend of the range, thus running into Arizona a few miles north of Rioville. In fact, the mica belt forms the boundary line between Nevada and Arizona for 50 miles The mica, mostly small, is abundant, but mar- ketable sizes are rare, and not to be had without a great deal of hard work. 5 Merchantable mica has been reported on the Payette River and Bear Creek, in the Coeur d'Alene region of Idaho, and also in Oregon and Alaska. According to Mr. R. W. Ells 6 the Canadian micas of commercial importance occur associated with eruptive dikes of pyroxenite and pegmatite cutting the Laurentian gneisses. More rarely, as in the Gatineau area, they are found where dikes of the pyroxenite cut the limestone. This authority gives the condition of occurrence as below: 1. In pyroxene intrusive rocks which either cut directly across the strike of grey- ish or other colored gneisses or are intruded along the line of stratification. Some of these deposits have been worked downward along the contact with the gneiss, where the mica is most generally found, for 250 feet, as at the Lake Girard Mine, and irregular masses of pink calcite are abundant. In certain places apatite crystals 1 Geology of the Black Hills of Dakota, Monograph, U. S. Geological Survey, 1880. 2 Engineering and Mining Journal, XXXVI, 1883, p. 145. 3 Transactions of the American Institute Mining Engineers, XVII, 1889, p. 570. Proceedings of the Boston Society of Natural History, XXIII, 1884-1888, p. 488. ;> Mineral Resources of the United States, 1893. p. 754. "Bulletin of the Geological Society of America, V, 1894, p. 484. 292 REPORT OF NATIONAL MUSEUM, 1899. occur associated with the mica, but at other times these are apparently wanting. As in the case of apatite deposits, mica occurring in this condition would apparently be found at almost any workable depth. 2. In pyroxene rocks near the contact of cross-dikes of diorite or feldspar, the action of which on the pyroxene has led to the formation of both mica and apatite. Numerous instances of this mode of occurrence are found, both in the mines of apatite and mica, the deposits of the latter in certain areas being quite extensive and the crystals of large size. 3. In pyroxene rock itself distinct from the contact with the gneiss. In these cases the mica crystals, often of large size but frequently crushed or broken, appar- ently follow certain lines of faults or fracture. Some of these deposits can be traced for several yards, but for the most part are pockety. Some of these pyroxene masses are very extensive, as in the case of the Cascade mine on the Gatineau river and elsewhere in the vicinity. In these cases calcite is rarely seen and apatite is almost entirely absent. When cut by cross-dikes conditions for the occurrence of mica or apatite should be very favorable. 4. Dikes of pyroxene, often large, cutting limestone through which subsequent dikes of diorite or feldspar have intruded as in Hincks township. The crystals occurring in the pyroxene near to the feldspar dikes are often of large size and of dark color, resembling in this respect a biotite mica. The mica found under the conditions stated above, in one, two, three, and four is all amber-colored and of the variety known as phlogopite, or magnesia mica. [Specimens Nos. 30763, 62149, U.S.N.M.] 5. In feldspathic-quartzose rocks which constitute dikes often of very large size, cutting red and greyish gneiss, as at Villeneuve and Venosta. These are distinct from the smaller veins of pegmatite which occur frequently in the gneiss as the anorthosite areas are approached. In this case the mica is muscovite or potash mica and is invariably found in that portion of the dike near the contact with the gneiss. The crystals frequently are of large size and white in color, associated with crystals of tourmaline, garnet, et cetera, but with no apatite, unless pyroxene is also present. 6. In quartz-feldspar dikes cutting crystalline limestone, in which case the crystals are generally of small size, mostly of dark color and of but little value. In the case of the amber micas this peculiarity was noted that when the pyroxene was of a light shade of greenish gray and comparatively soft, the mica was cor- respondingly light colored and clear, and in some places almost approached the mus- covite in general appearance. As the pyroxene became darker in color and harder in texture, the mica assumed a correspondingly darker tint and a brittle or harder character, and in certain cases where dikes of blackish hornblendic diorite were present the mica also assumes a black color as well. The chief Canadian localities, as given by the authority quoted, are as below: Along the Ottawa Eiver it is found from a point nearly 100 miles west of Ottawa to the township of Greenville, 60 miles east of that city, while on the Gatineau River, which flows into the Ottawa at the city of Ottawa, mines have been located and worked for 80 miles north from its mouth, and the mineral is reported from points many miles farther north along that stream. To the east of Quebec it is known on the branch of the Saguenay called the Manouan and in the townships of Escoumains, Bergeronnes, and Tadoussac, situated east of the mouth of that river, as well as at several other places along the river St. Lawrence. The mica found in this last district is chiefly muscovite. The principal areas where mica is at present worked are in the belt which extends from North Burgess, in the Province of Ontario, approximately along the strike of the gneiss, into the territory adjacent to the Gatineau and Lievre. Over much of this THE NONMETALLIC MINERALS. 293 area south of the Ottawa River the Lauren tian is concealed by the mantle of Cambro- Silurian rocks belonging to the Ottawa River basin, but it may be said that the geo- logic conditions and the stratigraphic sequence in the area south of the Ottawa and in the rear of Kingston are the same as those found in the mineral-bearing belts north of the Ottawa, and that the most favorable conditions under which the deposits of mica and apatite may be looked for w r here traces of igneous agency are visible in the presence of dikes of pyroxene and quartz feldspar, though it should be stated that the mere occurrence of these in the gneiss does not warrant the presence of either of these minerals. The India mica mines occur in coarse intrusive pegmatitic-granite dikes, cutting what is known as the "newer gneiss" of Singrauli. At Inikurti the crystals (of mica) are as much as 10 feet in diameter. Sheets 4 or 5 feet across have been obtained free from adventitious inclusions which would spoil their commercial value. 1 Black mica (biotite, lepidomelane, etc.,) is a much more common and widely distributed variety than the white, but unlike the latter is found not so much in veins as an original constituent disseminated in small flakes throughout the mass of eruptive and metamorphosed sedimentary rocks. The small sizes of the sheets, their color, and lack of transparency render the material as a rule of little value. In Renfrew County, Canada, the mineral occurs in large cleavable masses, which yield beautiful smoky-black and greenish sheets sufficiently elastic for industrial purposes (Specimens Nos. 62735, 62709, U.S.N.M.). Lepidolite. This variety of mica is much more rare than any of the others described. While in a few instances it has been reported as accompanying rnuscovite in certain granites, as those of Elba and Schaistausk, its common form of occurrence is in the coarse pegmatitic veins already described, where it is associated with muscovite, tourma- lines, and other minerals of similar habit. As a rule it is readily distin- guished from other micas by its beautiful peach-blossom red color, though sometimes colorless and to be distinguished only by the lithia reaction. 2 The folia are thicker than those of muscovite and of small size, the usual form being that of a scaly granular aggregate. At Au- burn, Maine, it is found both in this form (Specimen No. 61079, U.S.N.M.) and forming a border a half inch, more or less, in width about the muscovite folia (Specimen No. 13810, U.S.N.M.). The more noted localities in the United States are Auburn, Androscoggin County; Hebron, Paris, Rumford (Specimen No. 63003, U.S.N.M.), and Norway, Oxford County, Maine, where it is associated with beau- tiful red and green tourmalines and other interesting minerals; Ches- terfield, Massachusetts; Iladdam, Connecticut (Specimen No. 53540, U.S.N.M.), and near San Diego, California (Specimen No. 62593, U.S.N.M.). The most noted foreign locality is Zinnwald, Saxony, Geology of India, 2d ed., 1893, p. 34. 2 The pulverized mineral when moistened with hydrochloric acid and held on a wire in the flame of a lamp imparts to the flame a brilliant lithia red color. 294 REPORT OF NATIONAL MUSEUM, 1899. where the mineral occurs in large foliated masses together with quartz forming the gangue minerals of the tin veins. Also found in Moravia (Specimen No. 62580, U.S.N.M.). Uses, Until within a few years almost the only commercial use of mica was in the doors or windows of stoves and furnaces, the peep- holes of furnaces and similar situations where transparency and resist- ance to heat were essential qualities. To a less extent it was used in lanterns, and it is said, in the portholes of naval vessels, where the vibrations would demolish the less elastic glass. In early days it was used to some extent for window panes, and is, in isolated cases, still so used to some extent. For all these purposes the white variety musco- vite is most suited. For use in stoves and furnaces "the mica is gen- erally split into plates varying from about one-eighth to one sixty- fourth of an inch in thickness. In preparing these plates for market the first step is to cut them into suitable sizes. Women are frequently employed in this work, and do it as well as, if not better, than the men. The cutter sits on a special bench which is provided with a huge pair of shears, one leg of which is tirmly fixed to the bench itself, while the movable leg is within convenient grasp. The patterns according to which the mica is cut are arranged in a case near at hand. They are made of tin, wood, or pasteboard, accord- ing to the preference of the establishment. Generally they are simple rectangles, varying in size from about four square inches to eighty. The cutter selects the pattern which will cut to the best advantage, lays it on the sheet of mica, and then, holding the two firmly together, trims off the edges of the mica to make it correspond to the pattern. The cleaning process comes next. The cleaner sits directly in front of a window and must examine each sheet of cut mica by holding it up between her eyes and the light. If there be any imperfections, and there nearly always are, they must be removed by stripping off the offending layers of mica until a clear sheet remains. Finally, the cut and cleaned mica is put up in pound packages and is ready for the market. There is an enormous waste in the processes of preparation. One hundred pounds of block mica will scarcely yield more than about fifteen pounds of cut mica, and sometimes it is even less. The proportion varies, of course, with different localities. 1 Professor Kerr states with reference to the North Carolina mines that there is a waste of from nine-tenths to nineteen-twentieths of the material, even in a good mine. Mica being a nonconductor is of value for insulating purposes, and since the introduction of the present system of generating electricity there has arisen a considerable demand for it in the construction of dynamos and electric motors. For these purposes the mica must be 1 Engineering and Mining Journal, LV, 1893, p. 4. THE NONMETALLIC MINERALS. 295 smooth and flexible, as well as free from spots or inequalities of any kind. It is stated that it should be sufficiently fissile to split into sheets not above three one-hundredths inch in thickness, and which may be bent without cracking into a circle of 3 inches diameter. Strips of various dimensions are used in building up the armatures, the more common sizes being about 1 inch wide by 6 or 8 inches long. Musco- vite serves the purposes well, but is less used than phlogopite, the latter serving equally well, and being less desirable for stoves and fur- naces. Black mica would doubtless serve for electrical purposes, could it be procured in sheets of sufficient size. Mica scraps such as until within a few years have been thrown away as worthless are now utilized by grinding, the product being used for a variety of purposes, noted below. The material is as a rule ground to five sizes, such as will pass through sieves of 80, 100, 140, 160, and 200 meshes to the inch, respectively. The prices of this ground material vary from 5 to 10 cents a pound according to sizes. Large quanti- ties of this ground material are used in the manufacture of wall paper, in producing the frost effects on Christmas cards, in stage scenery, and. as a powder for the hair, being sold for the latter purposes under the name of diamond powder. The so-called French "silver molding" is said to be made from ground mica. It is also used as a lubricant, and as a nonconductor for steam and water heating; in the manufacture of door knobs and buttons. It is stated further that owing to its elas- ticity it can be used as an absorbent for nitroglycerin. rendering ex- plosion by percussion much less likely to occur. Small amounts of inferior qualities are also mixed with fertilizers where it is claimed to be efficacious in retaining moisture. A brilliant and unalterable mica paint is said to be prepared by first lightly igniting the ground mica and then boiling in hydrochloric acid, after which it is dried and mixed with collodion, and applied with a brush. Owing to the unalterable nature of the material under all ordinary conditions, and the fact that it can be readily colored and still retain its brilliancy and transparency, the ground mica is peculiarly fitted for many forms of decoration. Much of the ground material now produced is stated to be sent to France. The chief and indeed only use for lepidolite thus far developed is in the manufacture of the metal lithium and lithia salts. Prices. The total value of the cut mica produced annually in the United States during the past ten years has varied from $50,000 to over $360,000, while the value of the imports has varied between $5,000 and $100,000. The price of the cut mica, it should be stated, varies with the size of the sheets, the larger naturally bringing the higher price. The average price of the cut mica, all sizes, is not far from $1 a pound, while the scrap mica is worth perhaps half a cent a pound. The dealers' lists, as published, include 193 sizes, varying 296 REPORT OF NATIONAL MUSEUM, 1899. from 1| by 2 inches up to 8 by 10 inches, the prices ranging from 40 cents to $13 a pound. For electrical work upward of 400 patterns are called for, the prices varying from 10 cents to $2.50 a pound. 3. ASBESTOS. The name asbestos in its original sense includes only a fibrous variety of the mineral amphibole; hence is a normal metasilicate of calcium and magnesium with usually varying amounts of iron and manga- nese and not infrequently smaller quantities of the alkalies. As is well known, the amphiboles crystallize in the monoclinic system in forms varying from short, stout crystals, like common hornblende, to long columnar or even fibrous forms, to which the names actinolite, tremolite, and asbestos are applied. The word asbestos is derived from the Greek afffifffros, signifying incombustible, in allusion to its fireproof qualities. The name "amianthus" was given it by the Greeks and Romans, the word signifying undefiled, and was applied in allusion to the fact that cloth made from it could be readily cleansed by throwing it into the fire. As now used, this term is properly limited to fibrous varieties of serpentine. Owing to careless usage, and in part to ignorance, the name asbestos 1 is now applied to at least four distinct minerals, having in common only a fibrous structure and more or less fire and acid proof properties. These four minerals are: First, true asbestos; second, anthophyllite; third, fibrous serpentine (chryso- tile), and, fourth, crocidolite. The true asbestos is of a white or gray color, sometimes greenish or stained yellowish by iron oxide. Its fibrous structure is, however, its most marked characteristic, the entire mass of material as taken from the ledge or mine being susceptible of being shredded up into fine fibers sometimes several feet in length. In the better varieties the fibers are sufficiently elastic to permit of their being woven into cloth. Often, however, through the effect of weathering or other agencies, the fibers are brittle and suitable only for felts and other nonconducting materials. The shape of an asbestos fiber is as a rule polygonal in outline and of a quite uniform diameter, as shown in the illustration (fig. 10); often, however, the fibers are splinter like, running into fine, needle-like points at the extremity. The diameters of these fibers is quite variable, and, indeed, in many instances there seems no practical limit to the shredding. Down to a diameter of 0.002 mm. and sometimes to even 0.001 mm. the fibers retain their uniform diameter and polygonal outlines. Beyond this, however, they become splinter like and irregular as above noted. The mineral anthophyllite like amphibole occurs in both massive, platy, and fibrous forms, the fibrous form being to the unaided eye indistinguishable from the true asbestos. 1 Also spelled asbestus. The termination o seems most desirable when the deriva- tion of the word is considered. THE NONMETALLIC MINERALS. 297 Chemically this is a normal metasilicate of magnesia of the formula (Mg,Fe) SiO 3 , differing, it will be observed, from asbestos proper in containing no appreciable amount of lime. It further differs in crys- tallizing in the orthorhombic rather than the monoclinic system, a feature which is deterniinable only with the aid of a microscope. The shape and size of the fibers is essentially the same as true asbestos. The fibrous variety of serpentine to which the name asbestos is commer- cially given is a hydrated met- asilicate of magnesia of the formula H 4 Mg 3 Si 2 O 9 with usu- ally a part of the magnesia replaced by ferrous iron. It differs, it will be observed, from asbestos and anthophyl- lite in carrying nearly 14 per cent of combined water and from the first named in con- taining no lime. This mineral is in most cases readily distin- guished from either of the others by its soft, silk-like fibers and further by the fact that it is more or less decomposed by acids. As found in nature the material is of a lively oil yellow or greenish color, compact and quite hard, but may be readily reduced to the fluffy, fibrous state by beating, handpicking, or running between rollers. The length of the fiber is quite variable, rarely exceeding 6 inches, but of very smooth uniform diameter and great flexibility. The mineral crocidolite, although resembling somewhat fibrous ser- pentine, belongs properly to the amphibole group. Chemically it is anhydrous silicate of iron and soda, the iron existing in both the sesqui- oxide and protoxide states. More or less lime and magnesia may be present as combined impurities. The color varies from lavender blue to greenish, the fibers being silky like serpentine, but with a slightly harsh feeling. The composition of representative specimens of these minerals from various sources is given in the accompanying table. 1 'From Notes on Asbestos and Asbestiform Minerals by George P. Merrill. Pro- ceedings of the U. S. National Museum, XVIII, 1895, pp. 281-292. Fig. 10. ASBESTOS FIBERS. After G. P. Merrill. 298 REPORT OF NATIONAL MUSEUM, 1899. | S : 8 S; 3 8 g g 3 g S 8 85 o ^ 8 5-2- -S- ^ l^COCOiOOO (NO'-CCC^J 1 s- r~ I-H co o> o> loos-^c^^i S ?S?iSc5S SSSJSco ^ CO CS Ci Oi CO d o co *i S iS g 8 s g rl rH S S 8 8 5-5^ oooooi ciioo oo -o -o Q - . Q i-H ee I B i f 2 ! |1! 1 51 iHS C ^^ tT 3 2 S ^ Is -i H * Il S| ^ I 6 THE NONMEtALLlC MINERALS. 299 i & - I G 1 1 1 1 i 1 1 dddd^ 1 1 1 1 *; . g .5 .S .S .5 8 ^ a E K E W I 1 8 1 1 Hfil a o S s 1 1 ']> S S 1 r^ M ^H 10 05 Q C1 1C O 1C CO QC TP 1 888 8 S S SI g 8 SS S ? S S ' 8 S 8 S cj d ' S S ^ as s 1 o I SC4 iO iO OO t> iO CC rH 3 : : S s i 11 r.j [ ! .' I rH to ; ; ; S S d do" : i d : "* : : : S la S S d S : : _ ' : : ~^ S I si S 1 sl * s a - s a si a 8 S5 S S S g 8 S S 5 S S S e S P 1 Si!: CO CC c^ -^ ^H to c-J os o j2 iC r^ O 3 ) 8.39 8.67 5.60 5.26 2.84 7.68 5.06 11.61 5.39 Lanthanum sesquioxide (LaoO 3 ) 5.46 0.87 None. 2.92 4.11 None. Erbinum sesquioxide (Er 2 O 3 ) 0.52 7.86 None. 7.01 None. 9.89 Manganese (MnO) 1.66 10.48 0.64 12.78 Trace. 13.02 Magnesia (MgO) 0.08 28 0.11 40 1.11 02 Soda (NajO) None. 3 49 None. 9 37 0.28 2 56 99.07 100.79 99.19 (I) Hittero, Norway; (II) Ytterby, Sweden; (III) Nelson County, Virginia. When in crystals often in long slender nail-like forms (orthite); also massive and in embedded granules. Color pitch black, brownish, and yellow. Brittle. Hardness 5.5 to 6. Specific gravity 3.5 to 4.2. Before the blowpipe fuses and swells up to a dark, slaggy, magnetic glass. Localities and mode of occurrence. The more common occurrence is in the form of small acicular crystals as an original constituent in granitic rocks. It also occurs in white limestone, associated with mag- netic iron ore, and in igneous rocks as andesite, diorite, and rhyolite. At the Cook Iron Mines, near Port Henry, New York, it is reported as occurring in great abundance and in crystals of extraordinary size, in a gangue of quartz and orthoclase. The variety orthite occurs in forms closely simulating rusty nails in the granitic rock about Brunswick, Maine. In Arendal, Norway, it is found in massive forms (Specimen No. 66853, U.S.N.M.). At Finbo, near Falun, Sweden, in acicular ciystals a foot or more in length. In Amherst and Fauquier counties, Virginia, it occurs in large masses (Specimen No. 68661, U.S.N.M.) from Fauquier County, as it also does near Bethany Church, Iredell County, North Carolina, and Llano County, Texas (Specimen No. 62756, U.S.N.M.). At Balsam Gap, Buncombe County, North Carolina, it occurs in slender crystals 6 to 12 inches long and in crystalline masses, in a granitic vein and under similar conditions at the Buchanan and Wiseman mines in Mitchell County. Uses. See under Monazite, p. 383. THE NONMETALLIC MINERALS. 313 9. GADOLINITE. This is a basic orthosilicate of yttrium, iron, and glucinum, though with frequently varying- amounts of didymium, lanthanum, etc. The formula as given by Dana is Gl 2 FeY 2 Si 2 O 10 , silica 23.9 per cent, yttrium oxides 51.8 per cent, iron protoxide 14.3 per cent, and glu- cina 10 per cent. Actual analyses yielded results as below: Constituents. I. II. Silica (SiO 2 ) ''4 35 23 79 Thorina (ThO 2 ) 0.30 0.58 Yttrium sesquioxide (Y 2 Os) 45 % 41 55 Cerium sesquioxide (Ce 2 O 3 ) 1.65 2.62 Lanthanum sesquioxide (LaoO 3 ) Iron sesquioxide ( FeoO 3 ) } 3.06 2.03 5.22 0.% Iron protoxide (FeO) Berylium (Glucina) protoxide (BeO) ... Lime (CaO) 11.39 10.17 30 12. 42 11.33 74 Soda(NaoO) 17 Trace Water ( HO) 0.52 1.03 99.90 100.24 (I) Ytterby, near Stockholm, Sweden; (II) Llano County, Texas. The mineral is sometimes found in form of rough and coarse crystals, but more commonly in amorphous, glassy forms. Hardness 6.5 to 7; specific gravity 4 to 4.47. Color brown, black and greenish black, usually translucent in thin splinters and of a grass green to olive green color by transmitted light. No true cleavage; fracture conchoidal or splintery like glass, and with a vitreous or somewhat greasy luster. Through oxidation and hydration the mineral becomes opaque, brown, and earthy. Hence masses are not infrequently found consisting of the normal glassy gadolinite enveloped in a brown red crust of oxidation products. (Specimen No. 62780, U.S.N.M.) On casual inspection the mineral closely resembles samarskite and the dark, opaoue varie- ties of orthite, but is easily distinguished from the fact that before the blowpipe it glows brightly for a moment and then swells up, cracks open, and becomes greenish without fusing. Some varieties (the nor- mal anisotropic forms) swell up into cauliflower-like forms and fuse to a whitish mass. Like orthite, it gives a jelly when the powdered mineral is boiled in hydrochloric acid. Localities and mode of occurrence. The mineral occurs mainly in coarse pegmatitic veins associated with allanite, and other allied minerals. The principal locality in the United States thus far described is some five miles south of Bluffton on the west bank of the Colorado River, in Llano County, Texas (Specimen No. 62780, U.S.N.M). The region is described * as occupied by Archaean rocks with granite, and occasional cappings of limestone. 'American Journal of Science, XXXVIII, 1889, p. 474. 314 REPORT OF NATIONAL MUSEUM, 1899. A coarse deep red granite is the most abundant, and is cut by numer- ous extensive veins of quartz and feldspar which carry the gadolinite, in pockety masses, and the other minerals mentioned. Most of the mineral thus far found is altered into the brown-red waxy material noted above and occurs in the form of masses weighing half a pound and upward. One "huge pointed mass, in reality a crystal, weighed fully 60 pounds; " another 42 pounds. One of the earliest opened pockets yielded some 500 kilos (227 .pounds) of the mineral. Of the foreign localities those of Kararfvet, Broddbo and Finbo, near Falun, Sweden, and at Ytterby, near Stockholm (Specimen No. 62793, U.S.N.M.), are important, the mineral here occurring in the form of rounded masses embedded in a coarse granite. On the island of Hittero, in the Flecke fiord, southern Norway, crystals sometimes four inches across have been obtained. Uses. See under monazite, p. 383. 10. CERITE. This is a silicate of the metals of the cerium group; of a com- plex and doubtful formula. The analyses below, taken from Dana's System of Mineralogy, will serve to show the varying character of the mineral. Constituents. I. II. III. Silica (SiO) 19.18 22.79 18.18 Cerium oxide (CeO 3 ) 64 55 24 06 33 25 Didynium oxide (Di 2 O 3 ) Lanthanum (La^Os) }7. 28 35.37 34.60 Iron oxide (FeO) 1.54 3.92 3.18 Alumina (AU) 3 ) 1.26 Lime (CaO) 1.35 4.35 1.69 Water (H0) 5.71 3.44 5.18 The mineral occurs in gneiss and mica schist, and is of a prevailing pink to gray color. Specimen No. 62794, U.S.N.M., from Bastnass, Westmanland, Sweden, is characteristic, Uses. See under monazite, p. 383. 11. RHODONITE. This is a metasilicate of manganese of the formula MnSiO 3 , = Silica 45.9 per cent; manganese protoxide 54.1. As a rule, iron, calcium, or zinc replaces a part of the manganese. The prevailing form of the mineral when in crystals is that of rough, tabular, or elongated prisms with rounded edges (Specimen No. 83927, U.S.N.M., from Franklin, New Jersey). It is also common in massive highly cleavable forms, and in disseminated granules (Specimens Nos. 83927 and 83929, U.S.N.M.). Barely, as in the Ekaterinburg district of Russia, it occurs in massive THE NONMETALLIC MINERALS. 315 forms suitable for ornamental work. (See Collection Building and Ornamental Stones.) Color brownish red, flesh red, and pink; some- times rose red. Hardness, 5.5 to 6.5. Specific gravity, 3.4 to 3.68. On exposure the mineral undergoes oxidation, becoming coated with a black film and giving rise thus to indefinite admixtures of silicate, oxides, and carbonates of manganese. The mineral occurs in abundance associated with the iron ores of Wermland, Sweden, and at other localities in Europe; in Ekaterin- burg, Russia, as above noted. The zinciferous variety commonly asso- ciated with the zinc ores in granular limestones of Sussex County, New Jersey, is known as fowlerite. (Specimen No. 67405, U.S.N.M.) So far as the writer has information, rhodonite has as yet little com- mercial value, excepting as an ornamental stone. To some extent it has been utilized in glazing pottery and as a flux in smelting furnaces. 12. STEATITE; TALC; AND SOAPSTONE. The mineral steatite, or talc, is a soft micaceous mineral, consisting when pure of 63.5 percent of silica, 31.7 per cent of magnesia, and 4.8 per cent of water. Its most striking characteristics are its softness, which is such that it can be readily cut with a knife or even with the thumb nail, and soapy feeling, there being an entire absence of anything like grit. The prevailing colors are white or gray and apple green. Several varietal forms are recognized; the name talc as a rule being applied to the distinctly foliaceous or micaceous variety (Specimen No. 72838, U.S.N.M.), while that of steatite is reserved for the compact cryptocrystalline to coarsel v granular forms (Specimens Nos. 26137 and 63448, U.S.N.M.). Pyrallolite and rensselaerite are names given to varied forms of talc resulting from the alteration of hornblende or pyroxene. Such forms are found in various portions of northern New York, Canada, and Finland. According to. Dana, a part of the so-called agalmatolite used by the Chinese is steatite. The name soapstone is given to dark gray and greenish talcose rocks, which are soft enough to be readily cut with a knife, and which have a pronounced soapy or greasy feeling; hence the name. Such rocks are commonly stated in text-books to be compact forms of stea- tite, or talc, but as the writer has elsewhere pointed out, and as shown by the analyses here given, few of them are even approximately pure forms of this mineral, but all contain varying proportions of chlorite, mica, and tremolite, together with perhaps unaltered residuals of pyroxene, granules of iron ore, iron pyrites, quartz, and in seams and veins calcite and magnesian carbonates. 1 1 Rocks, Rock weathering, and Soils, p. 101. 316 REPORT OF NATIONAL MUSEUM, 1899. Composition. The varying composition of talc is shown in the series of analyses given below. Analyses of talc. Locality. SiO. A1 2 3 . St. Lawrence County, New York 60.59 0.13 Do 62.10 Luzenach, France 61.85 2.61 Valley of Pignerolles, Italy . 60.60 0.30 FeO. MgO. CaO. MnO. NaO. KO. 1 16 2.15 0. 17 Trace. .77777. 77777. 0.40 2.80 Totals. Not deter- min- ed. 100.00 100.00 The following analyses of soapstone have been made in the labora- tory of the department: Analyses of soapstone. Locality. SiO 2 . A1 2 3 . FeO. MgO. CaO. MnO. Na<>O. K 2 0. H 2 O. Totals. Francestown, New Hamp- shire (Specimen No. 63166, U SN M ) 42 43 6 08 13 07 25 71 3 27 16 32 8 45 99 49 Grafton, Vermont (Speci- men No. 17569, U.S.N.M.). 51.20 5.22 8.45 26.79 1.17 0.32 6.90 100.05 Dana, Massachusetts (Speci- men No. 26439, U.S.N.M.) . . 38.37 5.64 8.86 28.62 3 90 14 49 99 88 Baltimore County, Mary- land (Specimen No. 26628, U.S.N.M.) 52.70 5.57 7.63 1.77 5.48 100.03 Guilford County, North Car- olina (Specimen No. "7662, U.S.N.M.) 40.03 10.86 9.59 26.97 1.70 10.78 99.93 Lafayette, Pennsylvania (Specimen No. 63168, U.S.N.M.) 33.47 0.45 7.38 33.72 1.34 0.21 23.00 99 57 Occurrence and origin. Talc in all its forms is presumably always a secondary mineral, a product of alteration of other magnesian silicates. Smyth has shown * that the talc beds of St. Lawrence County, New York (Specimen No. 63173), are alteration products from schistose aggregates of enstatite or tremolite, principally the former. Accord- ing to this author, the talc occurs, not as has been stated, in the form of a well-defined vein with walls of granite or gneiss, but in the beds lying wholly within the schistose portions of the prevailing limestone. The following account of these deposits as occurring near Gouv- erneur is by A. Sahlin : 2 The village of Gouverneur is situated near the northwest edge of a geological island of Azoic rocks; granite, gneiss, limestone, and marble School of Mining and Forestry, XVII, No. 4, 1896. Also Fifteenth Annual Report of the State Geologist of New York, 1895, pp. 665-671. 2 Mining and Scientific Press, May 11, 1893. THE NONMETALLIC MINERALS. 317 being the representative features of the formation. To the west of Gouverneur, extending to and beyond the St. Lawrence River, the Potsdam sandstone is encountered; to the southeast, the Trenton lime- stones extend toward the Adirondack Mountains. The talc belt is found in the towns of Fowler and Edwards, from 7 to 14 miles south- east of Gouverneur. It has a length of about 8 miles, a width of 1 mile, more or less, and crosses the above-named Azoic island in the general direction of WNW. to ESE. The "veins" generally dip from 45 to 75 toward the northeast. Their width varies from a few inches to 20 feet or more. Surface out croppings are frequent, and local experts contend that there is no use in looking for talc where it does not appear on the surface. The abrupt change of formation precludes the prob- ability of discovering new deposits beyond the small, and now most thoroughly explored, belt already known. Within this narrow terri- tory, "veins" of talc minerals, separated by layers of granite and gneiss, are found and worked. They are principally made up of the hydrated silicates of magnesia, known as agalite and rensselaerite, the former of a smooth, fibrous texture, the latter scaly and lamellar, and both beautifully white or bluish-white. In the agalite veins are found nodules of handsome pink to purple, columnar crystals of hexagonite, and also large "horses" of yellowish- white hornblende. The occur- rence of the two latter minerals, representing the anhydrous silicates of magnesia, has given rise to the theory that the talc deposits origi- nally occurred as hornblende, which has gradually become hydrated. Since 1879, ten distinct mines have been opened, and some of these have reached a depth of 400 feet or more on the slope. The present output from these ten mines amounts, according to a close estimate, to 51,000 tons a year, which figure, however, could be readily doubled if the reducing mills had the capacity to handle the larger quantity. (Specimens Nos. 53590 to 53592, U.S.N.M., from Gouverneur are characteristic.) In western North Carolina and northern Georgia, particularly in Cherokee, Moore, Guilford, and Murphy counties in the first-named State, and in the Cohutta Mountains of Murray County in the last, are numerous beds of very clean white or greenish fibrous talc occurring in part, at least, in connection with the marble beds. Some of the material is soft, white, and almost translucent (Specimens Nos. 26137, 27654, 63448, U.S.N.M.), while other is tough and semitranslucent, hornlike. The beds are mostly very irregular in extent as well as in quality of material. In Stockbridge, Windsor County, Vermont, talc is mined from veins from 3 to 12 feet in width in soapstone. (Specimen No. 53206, U.S.N.M.) A greenish schistose talc is also mined in Murray County, Georgia, (Specimen No. 53226, U.S.N.M.) Soapstone occurs mainly associated with the older crystalline rocks, 318 REPORT OF NATIONAL MUSEUM, 1899. and in some eases is undoubtedly an altered eruptive; in others there is a possibility of its being a product of metamorphism of magnesian sediments. The principal beds now known lie in the Appalachian regions of the eastern United States, though others have recently been found in California, and there is no reason for supposing that many more may not exist in the Rocky Mountain regions. The beds, if such they can be called, are not extensive as a rule, but occur in lenticular masses of uncertain age intercalated with other magnesian and horn- blendic or micaceous rocks frequently more or less admixed with ser- pentine. The rock, like serpentine, is, as a rule, traversed by bad seams and joints, and the opening of any new deposit is always attended with more or less risk, as there is in many cases no guarantee that sound blocks of sufficient size to be of value will ever be obtainable. The following facts relative to the occurrence of soapstone in the United States are taken mainly from a handbook by the writer on Stones for Building and Decoration, issued by Messrs. Wiley & Co., of New York. An extensive bed of fine quality soapstone was discovered as early as 1794 at Francestown, New Hampshire (Specimen No. 10774, U.S.N.M.). This was worked as early as 1802, and up to 1867 some 5,500 tons had been quarried and sold. In this latter year some 3,700 stoves were manufactured by one company alone. The business has been conducted on a large scale ever since, and the bed has been followed some 400 feet, the present opening being 40 feet wide 80 feet long and 80 feet deep. Other beds, constituting a part of the same formation, occur in Weare, Warner, Canterbury, and Richmond, in the same State, and all of which have been operated to a greater or less extent. Fine beds of the stone also occur in the town of Orford, and an important quarry was opened as early as 1855 in Haverhill, but it has not been worked continuously. At least sixty beds of soapstone are stated to occur in Vermont, mostly located along the east side of the Green Mountain range, and extending nearly the entire length of the State. The rock occurs asso- ciated with serpentine and hornblende, and the beds as a rule are not continuous for any distance, but have a great thickness in comparison with their length. It not infrequently happens that several isolated outcrops occur on the same line of strata, sometimes several miles apart, and in many cases alternating with beds of dolomitic limestone that are scattered along with them. The sixty beds above mentioned occur mainly in the towns of Reads- boro, Marlboro, New Fane, Windham (Specimen No. 26626, U.S.N.M.), Townsend, Athens, Grafton, Andover, Chester (Specimen No. 53244, U.S.N.M.), Cavendish, Baltimore. Ludlow, Plymouth, Bridgewater, Thetford, Bethel, Rochester, Warren, Braintree, Waitsfield, Moretown, Duxbury, Waterbury, Bolton, Stow, Cambridge, Waterville, Berk- shire, Eden, Lowell, Belvidere, Johnson, Enosburg, Westfield, Rich- Report of U. S. National Museum, 1899,-Merrill. PLATE 14. THE NONMETALLIC MINERALS. 319 ford. Troy, and Jay. Of these beds those of Grafton (Specimen No. 17569, U.S.N.M.) and Athens are stated to have been longest worked and to have produced the most stone. The beds lie in gneiss, and were profitably worked as early as 1820. Another important bed occurs in the town of Weatherfield. This, like that of Grafton, is situated in gneiss, but has no overlying rock, and the material can be had in inex- haustible quantities. It was first worked about 1847. The Rochester beds were also of great importance, the stone being peculiarly fine- grained and compact. It was formerly much used in the manufacture of refrigerators. The bed at New Fane occurs in connection with ser- pentine, and is some half mile in length by not less than 12 rods in width at its northern extremity. The soapstone and serpentine are strangely mixed, the general courses of the bed being like that of an irregular vein of granite in limestone. In Massachusetts quarries of soapstone have been worked from time to time in Lynnfield and North Dana (Specimen No. 26439, U.S.N.M.). The Lynnfield stone occurs associated with serpentine. It has not been quarried of late, but was formerly used for stove backs, sills, and steps. In New York State soapstone and talc occur in abundance near Fowler and Edwards in St. Lawrence County. Some of this is very pure, nearly snow-white talc, and is quarried and pulverized for commercial purposes, as already noted. In Pennsylvania, in the southern edge of Montgomery County, extending from the northern brow of Chestnut Hill between the two turnpikes across the Wissahickon Creek and the Schuylkill to a point about a mile west of Marion Square, there occurs a long, straight out- crop of steatite and serpentine. The eastern and central part of this belt on the southern side consists chiefly of steatite, while the northern side contains much serpentine, interspersed through it in lumps. Only in a few neighborhoods, as at Lafayette, does either the steatite or serpentine occur in a state of sufficient purity to be profitably quarried. On the east bank of the Schuylkill, about 2 miles below Spring Mill, a good quality of material occurs that has long been successfully worked (Specimen No. 63168, U.S.N.M.) The material is now used principally for stoves, fireplaces, and furnaces, though toward the end of the last century and during the early part of the present one, before the introduction of the Montgomery County mar- ble, it was in considerable demand for doorsteps and sills. It proved poorly adapted for this purpose, owing to the unequal hardness of the different constituents, the soapstone wearing away rapidly, while the serpentine was left projecting like knots, or " hobnails in a plank." Several small deposits of soapstone occur in Maryland and some of them have been worked on a small scale. The material is of good quality, but apparently to be had only in small pieces (Specimens Nos. 25010 and 26628) from Montgomery and Baltimore Counties. In Virginia soapstone occurs in Fairfax (Specimens Nos. 25254, 28649, 320 REPORT OF NATIONAL MUSEUM, 1899. U.S.N.M.), Fluvanna and Buckingham, counties. There is also a bed at Alberene, Albemarle County, a little west of Green Mountain. This is the bed so extensively worked by the Albemarle Soapstone Com- pany (Specimen No. 62547, U.S.N.M.) From these points the beds extend in a southwesterly direction through Nelson County, where they are associated with serpentine; thence across the James River above Lynchburg and present an outcrop about 2 miles west of the town on the road leading to Liberty; also one some 2 miles west of New London. Continuing in the same direction the bed is seen at the meadows of Goose Creek, where it has been quarried to some extent. Parallel ranges of soapstone appear near the Pigg River in Franklin County. About 30 miles southwest from Rich- mond, at Chula, in Amelia County, there are outcrops of soapstone said to be of fine quality, and which in former times were quite extensively operated by the Indians. They have been reopened within a few years and the material is now on the market. North Carolina contains, in addition to an abundance of the finest grades of talc and steatite as already noted, beds of the compact com- mon soapstone. Deposits in Cherokee and Moore counties furnish especially desirable material for lubricating and other purposes. Murphy, Guilford, Ashe, and Alamance counties (Specimen No. 27664, U.S.N.M.) are also capable of affording good materials, but much of it is inaccessible at present on account of poor railroad facilities (Specimens Nos. 27662, 28118, U.S.N.M.). from Greensboro and Ball Mountain. Beds of soapstone are stated to occur in Salina County, Arkansas (Specimen No. 39061, U.S.N.M.), and in Chester, Spartanburg, Union, Pickens,Oconee, Anderson, Abbeville, Kershaw, Fail-field, and Richland counties in South Carolina (Specimens Nos. 37590, 39019, U.S.N.M.). Texas is also stated to have an abundance of material and of good quality on the Hondo and Sandy creeks in Llano County. The Dis- trict of Columbia contains a bed which is, however, probably too small to ever prove of value (Specimen No. 38510, U.S.N.M.). Uses. The use to which the material is put varies greatly according to its purity and physical characteristics. The white, fibrous variety of great purity from St. Lawrence County, New York, is used as a filler in paper manufacture, something like 30 per cent of the weight of printing, paper being made up of this material. For the purpose it is run successively through coarse and finer crushers and then through buhrstones, after which it is placed into what is known as an Alsing cylinder, some 6 feet in diameter by about the same length. This cylinder is lined with porcelain brick and filled to one-third its volume with rounded pebbles or quartz, and when in motion revolves at about the rate of 20 revolutions a minute. At the end of some three to four hours the talc is reduced to the form of an impalpable powder. The so-called cyclone crusher has also been used to good advantage in this THE NONMETALLIC MINERALS. 321 work. The pulverized material is also used as a lubricator, for which purposes it is remarkably well adapted. Rubbed between the thumb and finger the powder is smooth and oily without a particle of grit. It is also used in soap making, for which purpose it can, however, be considered only as an adulterant, increasing the weight but not the cleaning properties of the article. It is further used as a dressing for fine leathers. Small quantities are used by shoe and glove dealers also. The pure, creamy white talc, such as is obtained from North Carolina, is used for crayons and slate pencils, while the still finer, cryptocrys- talline varieties, such as are at present obtained almost wholly from abroad, are used by tailors under the name of "French chalk" and for making the tips for gas burners. Fine compact grades of a some- what similar rock (agalmatolite) are used extensively in China and Japan for small ornaments. The stone is readily carved in fine sharp lines, and is a general favorite for making the grotesque images for which these countries are noted, and which are often sold throughout the country under the name of jadestone. The following account of the soapstone industry of China is taken from the Engineering and Mining Journal of September 30, 1893. The material referred to as soapstone is, however, very probably agalmato- lite. (See p. 322.) The British consul at Wenchow, in his last report, gives some interesting details respecting the manufacture of steatite or soapstone ornaments in China. The mines are distant 42 miles from Whenchow, and are reached by a boat journey of 35 miles up the river, followed by a land journey of 7 miles over rough ground. The hills containing steatite are owned by 20 to 30 families, who in some cases work the mines themselves, in others engage miners to do it on their account. The gal- leries are driven into the sides of the hills, and are often nearly a mile in length. The composition of the hills is soft, and the shafts require to be propped up by sup- ports of timber; for the same reason the floors are full of mire and clay, so that the miners wear special clothing, made principally of rhea fiber. They lead a hard life, living in straw huts on the hillside. The stone when first extracted is soft, hardening on exposure to the air. It is brought out of the mine in shovels, and is sold at the pit mouth to the carvers at a uniform price of about one-half a penny per pound. This would be when the purchaser buys it in gross, without first selecting it in any way. When picked over, the mineral varies very considerably in value according to the size of the lump, its shape, and above all, its colors. The colors are given as purple, red, mottled red, black, dark blue, light blue, gray, white, eggshell white, "jade," beeswax, and "frozen." Of these "jade" (the white variety, not the green) and "frozen" are the most valuable. Indeed so valuable is the latter that good speci- mens of it are said to fetch more than real jade itself. The industry finds employ- ment at the present time for some 2,000 miners and carvers. A great impetus was given to it by the opening of Wenchow to foreign trade. Previous to that event the chief purchasers of soapstone were officials and literary men, and the article most often carved was a stamp or seal. When it was discovered that foreigners admired the stone, articles were produced to meet what was supposed to be their taste. Such were landscapes in low or high relief, flower vases, plates, card trays, fruit dishes, cups, teapots, and pagodas. If left to his own devices, the native carver proceeds first to examine his stone, much as a cameo cutter would do, to discover how best he can take advantage of its shape and shades of color. ( See further under Agalmatolite. ) NAT MUS 99 21 322 REPORT OF NATIONAL MUSEUM, 1899. The following quotation from an English writer will serve to show the advantages gained by a use of talc in paper making: There is a decided advantage in substituting agalite for China clay, because not only is there an increase of dry paper, but such is obtained by a saving of fiber, as well as a decrease of the waste in the actual loading material and a lessened amount of polluting matter to be dealt with. Moreover, the fibrous character of the agalite causes it to yield a paper of higher class quality than is the case with China clay. The extra gloss which it is possible to obtain with papers containing agalite is shown in various American journals and books. The soapstones are suited for a considerable range of application. Although so soft, they are among the most indestructible and lasting of rocks, but are too slippery and perhaps of too sombre a color for general structural purposes. At present the chief use of the material in the United States is in the form of thin slabs for sinks and stationary washtubs. At one time it was quite extensively used throughout New England in the manufacture of stoves for heating purposes and to some extent for fire brick, the well-seasoned stone being thoroughly fire- proof. The putting upon the market of unseasoned materials or of material with bad veins, which caused the stone to crack or perhaps fly to fragments when subjected to high temperature, aroused a preju- dice against the employment of this material, and the manufacture is stated to have been to a considerable extent discontinued as a conse- quence. In the manufacture of either stoves or washtubs slabs of considerable size, free from segregation nodules of. quartz, pyrite, or other minerals or from dry seams, are essential. As but few of the now known outcrops can furnish material of this nature, the main part of the business of the country is in the hands of but two or three companies. The waste material from the quarries, or the entire out- put in certain cases, is pulverized and used as a lubricant or white earth, as is the micaceous variety. 13. PYROPHYLLITE; AGALMATOLITE; AND PAGODITE (IN PART). This is a hydrous silicate of aluminum corresponding to the formula H 2 O, A1 2 O 3 , 4SiO 2 . The analyses given below show the average com- position of the material as it occurs in nature: Locality. Silica. Aluminum. Water. Remarks. Westana, Sweden China 65.61 66 38 26.09 7.08 With small amounts of Deep River, North Carolina . . . 65.93 29.54 5.40 lime. The mineral is not known in distinct crystals, but occurs rather in foliated lamellar, massive and compact forms, closely resembling some forms of talc, for which its soapy or greasy feeling renders it very likely to be mistaken, though its hardness (2 to 2.5) is somewhat THE NONMETALLIC MINEEALS. 323 greater. The prevailing colors are white or greenish gray to dull red, often mottled. Occurrence. The material sometimes occurs, as in the Deep River region (Chatham, Moore, and Orange counties), North Carolina, in com- pact to schistose masses of beds of considerable extent and purity. Uses.- The more compact varieties, like that of Deep River (Speci- men No. 27665, U.S.N.M.), are used for making slate pencils and tailors' chalk, or French chalk, so called. The still more compact forms, known as agalmatolite (Specimens Nos. 37812, from Sonora, Mexico, and 27133 and 27134, Japan) and pagodite, are used extensively by the Chinese and Japanese for making small images and art objects of various kinds. Dana states, however, that a part of the so-called Chinese agalmatolite is in reality pinite and a part of steatite. The objects sold by Chinese dealers at the various expositions of late years under the name of jade stone are, however, of agalmatolite. FINITE: Agalmatolite in part. Composition, a hydrous silicate of alumina and the alkalies. According to Dana, 1 the name is made to include a large number of alteration products of white spodumene, nepheline, feldspar, etc. Professor Heddle has described 2 a pinite (agalmatolite) occurring in large lumps of a sea-green color, surround- ing crystalline masses of feldspar in the granites of Scotland, and which he regards as alteration products of oligoclase. The composition as given is: Silica, 48.72 per cent; alumina, 31.56 per cent; ferric oxide, 2.43 per cent; magnesia, 1.81 per cent; potash, 9.48 per cent; soda, 0.31 per cent; water, 5.75 per cent. 14. SEPIOLITE; MEERSCHAUM. This mineral is a hydrous silicate of magnesia, having the composi- tion indicated by the formula H 4 Mg 2 Si 3 O 10 , = silica 60.8 per cent; mag- nesia, 27.1 per cent; water, 12.1 per cent. The prevailing colors are white or grayish, sometimes with a faint yellowish, reddish, or bluish green tinge. It is sufficiently soft to be impressed by the nail, opaque, with a compact structure, smooth feel, and somewhat clay-like aspect; rarely it shows a fibrous structure. Specimens Nos. 62545, 66861, and 67749 are characteristic. In nature it rarely occurs in a state of absolute purity. The following analyses are quoted from Dana's Mineralogy : Locality. SiO 2 . MgO. FeO. HO. C0 2 . 61 17 28 43 06 9 83 67 Greece 61.30 28.39 0.08 9 74 0.56 Utah (fibrous) 52.97 22.50 r CuO. } 9.90 i Hygroscopic H 2 .O ' 1 System of Mineralogy, 6th ed., p. 621. 2 Mineralogical Magazine, IV, p. 215. 324 REPORT OF NATIONAL MUSEUM, 1899. The name is from the German words Meer, sea, and Schaum, foam, in allusion to its appearance. Mode of occurrence and origin. According to J. Lawrence Smith, 1 the Asiatic material occurs in the form of nodular masses in alluvial deposits on the plain of Eski-Shehr, and is regarded by him as derived by a process of substitution from magnesium carbonate which is found in the serpentine of the neighboring mountains. In an article by Dr. E. D. Clarke in the Cyclopedia of Arts and Sciences it is stated that the meerschaum of the Crimeria forms a stratum some 2 feet thick beneath a much thicker stratum of marl. Cleveland in his elementary treatise on minerals (1822) states that at Analotia, in Asia Minor, meerschaum occurs in the form of a vein more than 6 feet wide, in compact limestone. At Vallecas, Spain, a very impure form is stated to occur in the form of beds and in such abundance as to be utilized for building material. Aside from the localities above mentioned, sepiolite is known to occur in Greece, at Hrubschitz in Moravia, and in Morocco, in all cases being associated with serpentine, with which it is apparently genetically related. Uses. The mineral owes its chief value to its adaptability for smokers' use, being utilized in the manufacture of what are known as meerschaum pipes. At Vallecas, as above noted, the material is said to occur in such abundance as to be utilized as a building stone. In Algeria a soft variety is used in place of soap at the Moorish baths and for washing linen. According to Kunz, 2 meerschaum has occasionally been met with in compact masses of smooth, earthy texture in the serpentine quarries of West Nottingham Township, Chester County, Pennsylvania. Only a few pieces were found, but they were of good quality. It also occurs in grayish and yellowish masses in the serpentine in Concord, Dela- ware County, Pennsylvania. Masses of pure white material, weighing a pound each, have been found in Middletown, in the same county, and of equally good quality at the Cheever Iron Mine, Richmond, Mas- sachusetts, in pieces over an inch across; also in serpentine at New Rochelle, Westchester County, New York. A fibrous variety, in masses of considerable size, has within a few years been found in the Upper Gila River region, New Mexico (Specimen No. 67840, U.S.N.M.). According to a writer in the Engineering and Mining Journal, 3 the Eski-Shehr mineral is mined from pits and horizontal galleries in much the same manner as coal. As first brought to the surface it is white, with a yellowish tint, and is covered with red clayey soil. In this condition it is sold to dealers on the spot. Before exporting the 1 American Journal of Science 1849, VIII, p. 285. 2 Gems and Precious Stones, p. 189. 3 Volume LIX, 1895, p. 464. TH.E NONMETALLIC MINERALS. 325 material is cleaned, dried, and assorted, the drying taking place in the open air, without artificial heat in summer, and requiring from five to six days. The bulk of the material is sent direct to Vienna and Paris. 15. CLAYS. The term "clay," as commonly used comprises materials of widely diverse origin and mineral and chemical composition, but which have in common the property of plasticity when wet, and usually that of becoming indurated when dried either by natural or artificial means. Of so variable a nature is the material thus classed that no brief definition can be given that is at all satisfactory. One may perhaps describe the clays, as a whole, as heterogeneous aggregates of hydrous and anhydrous aluminous silicates, free quartz, and ever-varying quan- tities of free iron oxides and calcium and magnesian carbonates, all in a finely comminuted condition. Origin and mode of occurrence. The clays are invariably of sec- ondary origin that is, they result from the decomposition of pre- existing rocks and the accumulation of their less soluble residues, either in place (as residual clays) or through the transporting power of ice and water (drift clays). The fact that silicate of aluminum is so char- acteristic a constituent of nearly all clays is due to the fact that this substance is one of the most insoluble of natural compounds, and hence when, under the action of atmospheric or subterranean agencies, rocks decompose and their more soluble constituents as lime, mag- nesia, potash, soda, or even silica are removed, the aluminous silicate remains. The kaolins, which may perhaps be regarded as the simplest of clays, are the product, as a rule, of decomposition in place of feldspathic rocks, as gneisses, granites, and pegmatites. Those of Hockessin, Delaware (Specimens Nos. 63427 to 63430), are mainly of gneissic origin, though from some of the pits the material is in part at least derived from the decomposition of feldspathic conglomerate. In other cases the rock, as in the case of that from Blandford, Massachusetts (Speci- mens Nos. 68219 and 68221, U.S.N.M.), is a quite pure pegmatite, com- posed almost wholly of quartz and orthoclase. The samples show the material in various stages of decomposition. In all these cases the material as mined contains particles of free quartz and other substances detrimental to its use as a clay, and which must be removed by washing. It sometimes happens that the natural admixture of silica and unde- composed silicates is of just the right proportions to be utilized after merely griixling and bolting. The so-called "Cornwall stone" (Speci- mens Nos. 65136 and 62118, U.S.N.M.) is but a granite, very free from mica and ferruginous impurities, and in which the feldspar only has in part decomposed to the condition of kaolin. In some instances the natural conditions are such that running waters have assorted out the 326 REPORT OF NATIONAL MUSEUM, 1899. fine clay particles from the coarser jmpurities and deposited them by themselves, as in the case of that from Florida (Specimen No. 67256, U.S.N.M.). In the majority of cases, however, natural washing- has but served to still further contaminate the materials, giving rise to the complex transported clays to be noted later. Many rocks, such as the aluminous limestones, are so impure that on decomposing and the losing of their soluble lime carbonates they leave only very inferior varieties of clay, suitable for brick and tile or pottery making. Such are often highly colored by iron oxides (Specimens Nos. 62564, 62673, 63463, and 63493, U.S.N.M., in Rockweathering series). The assorting and transporting power of running waters rarely allow the beds of kaolin or of clay to remain in a condition of virgin purity or even in the place of their origin. The minute size and the shape of their constituent particles render them easily transported by rains and running streams, to be deposited again in regularly laminated beds (see Plate 18) when the streams lose their carrying power by flowing into lakes or seas. It is through such agencies that have in times passed been formed the so-called Leda clays (Specimen No. 73036, U.S.N.M.) and the loess. Such may contain a very large proportion of mechanically derived material and proportionately little kaolin. Speaking of clays of this nature as they exist in Wisconsin, Cham- berlain says: They owe their origin mainly to the mechanical grinding of glacial ice upon strata of limestone, sandstone, and shale, resulting in a comminuted product that now contains from 25 to 50 per cent of carbonates of lime and magnesia. This product of glacial grinding was separated from the mixed stony clays produced by the same action by water either immediately upon its formation or in the lacustrine epoch closely following. The process of separation must have been rapid and comparatively free from the agency of carbonated waters, otherwise the lime and magnesia would have been leached out. The formation of beds of clay has been confined to no particular period of the earth's history, but has evidently gone on ever since the first rocks were formed and when rock decomposition began. The older beds are as a rule greatly indurated and otherwise altered, and in many instances no longer recognizable as clays at all. Throughout the Appalachian region clay beds of Cambrian and Silurian ages have, by the squeezing and sheering incident to the elevation of this mountain system, become converted into argillites and roofing slates. Mineral and chemical composition. Formed thus in a variety of ways, and consisting not infrequently of materials brought from diverse sources, it is easy to comprehend that the substances ordinarily grouped under the name of clays may vary widely in both mineral and chemical composition. It may be said at the outset that the statements so fre- quently made to the effect that kaolinite or even kaolin is the basis of of all clays is not yet well substantiated. TflE NONMETALLIC MINEEAL8. Kaolinite is in itself not properly a clay, nor is it plastic. Further, in many cases it is present only in nonessential quantities. More open to criticism yet, because more concise, is the statement sometimes made that clay is a hydrated silicate of alumina having the formula A1 2 O 3 , 2SiO 2 +2H 2 O. It is doubtful if a clay was ever found which could be reduced to such a formula excepting by a liberal exercise of the imagi- nation. There is scarcely one of the silicate minerals that will not when sufficiently finely comminuted yield a substance possessing those peculiar physical properties of unctuous feel, plasticity, and color, which are the only constant characteristics of the multitudinous and heterogeneous compounds known as clays. Even pure vitreous quartz when rubbed to the condition of an impalpable powder has when wet the plasticity and odor of clay. 1 Daubree so long ago as 1878 2 pointed out the fact that by the mechanical trituration of feldspars in a revolv- ing cylinder with water an impalpable mud was obtained, which remained many days in suspension, and on drying formsd masses so hard as to be broken only with a hammer, resembling the argillites of the coal measures. The ever varying chemical nature of the materials classed as clays is brought out to some extent by a comparison of the analyses in the table (p. 349), but is even more evident in microscopic and mechanical examinations. Indeed, as stated by Chamberlain: 3 While it is convenient and customary to speak of the crude material of brick as clay, that which is really made use of is a mixture of clay and sand, or, in the cream- colored brick, of aluminous clay, calcareous clay or marl, and sand. The mixture is really a loam and but for the appropriation of that term as the designation of a soil, it would doubtless be more generally applied to such mixtures. Professor Crosby, as noted elsewhere, has shown that the blue-gray brick clays of Cambridge contain only from one-fourth to one-third of their bulk of "true clay," the remainder being finely comminuted material to which he gives the name rock flour. An examination of certain English fire clays has shown 4 that they can not properly be considered as mere hydrous silicates of alumina, but are very complex mineral admixtures, among which scales of hydrous micas, grains of feldspar, more rarely quartz and rutile needles greatly preponderate over the kaolin. The Leda clays of Maine, as the writer has noted elsewhere, contain a comparatively small amount 1 Referring to the odor of clay when a shower of rain first begins to wet a dry, clayey soil, Mr. C. Tomlinson has remarked that it is commonly attributed to alumina, and yet pure alumina gives off no odor when breathed upon or wetted. The fact is, the peculiar odor referred to belongs only to impure clays, and chiefly to those that con- tain oxide of iron. (Proceedings of the Geological Association, I, p. 242; quoted in Woodward's Geology of England and Wales, p. 439.) 2 Geologic Experimentale; 1879, p. 251. 3 Geology of Wisconsin, I, p. 673. *W. M. Hutchings, Geological Magazine, VII, 1890, p. 264, and VIII, 1891, p. 164. 328 REPORT OF NATIONAL MUSEUM, 1899. of kaolin but much free quartz, scales of mica, bits of still fresh feld- spar, and more rarely tourmalines and other of the less destructible silicates. Iron in the hydrated sesquioxide state is found in nearly all clays, even the whitest varieties. More than 1 per cent was found in a sili- ceous clay from Ohio, although the clay itself was almost of snowy whiteness. Iron also exists in the form of a silicate and protoxide carbonate, and sometimes as a sulphide in the form of disseminated pyrite. Lime and magnesia are also common constituents, either as free carbonates or as lime-magnesia silicates, and may exercise an important bearing upon the suitability of a clay for any particular purpose, as will be noted later. The clay from which the well-known Milwaukee cream- colored bricks are made contains sometimes as high as 23 per cent carbonate of lime and 17 per cent carbonate of magnesia, together with nearly 5 per cent of iron. The alkalies, potash and soda, are common constituents in small pro- portions, and also lithia, the first named being most common as well as most detrimental. It is a fair assumption that these substances are constituent of still undecomposed fragments of feldspar and the micas. To the presence of rutile needles and particles of ilmenite are due the frequent traces of titanic acid revealed by chemical analysis. The presence of any quartz and undecomposed feldspathic material in a clay can as a rule be detected by the gritty feeling manifested when tho material is rubbed between the thumb and fingers. Mica is, how- ever, not readily detected by this means. The above remarks will explain why a purely chemical analysis of a clay may be of little value for the purpose of ascertaining its suitability for any particular purpose. It is essential that we know not merely the presence or absence of certain elements but also how these elements are combined. Further than this few clays are used in their natural condition, being first purified by washing and usually mixed with other constituents to give them body or fire-resisting properties. Kinds and classification. From a geological standpoint the clays may be divided into two general classes, as above noted, (1) residual and (2) transported, the first class including a majority of the kaolin, halloysite, etc., and the second the ordinary brick and potter's clays, the loess, adobe, Leda, and the bedded, alluvial deposits of the Cre- taceous, Carboniferous, and other geological periods. Special names, based upon such properties as render them peculiarly adapted to eco- nomic purposes, are common. We thus have (1) the kaolin and China clay, (2) potter's clay, (3) pipe clay, (4) fire clay, (5) brick, tile, and terra cotta clays, etc., (6) slip clays, (7) adobe, and (8) fuller's earth. These will be discussed in the order given, though they must necessarily be discussed but briefly, since the subject of clays alone THE NONMETALLIC MINERALS. 329 could be made to far exceed the entire limits of the present volume. The names fat and lean clays are workmen's terms for clays relatively pure and plastic or carrying a large amount of mechanical admixtures, such as quartz sand. In the Kaolin and China Clays are included a series of clays used in the manufacture of the finer grades of porcelain and china ware and which consist in large proportion of the material kaolin, the name being derived from the Chinese locality Kaoling, from whence have for ages been obtained the materials for the highest grades of Chinese porcelain. According to Richthofen, 1 however, the material from which the porcelain of King-te-chin is made is not kaolin at all, but a hard greenish rock having somewhat the appearance of jade and which occurs intercalated between beds of clay slate. He says: This rock is reduced, by stamping, to a white powder, of which the finest portion is ingeniously and repeatedly separated. This is then moulded into small bricks. The Chinese distinguish chiefly two kinds of this material. Either of them is sold in King-te-chin in the shape of bricks, and as either is a white earth, they offer no visible differences. They are made at different places, in the manner described, by pounding hard rock, but the aspect of the rock is nearly alike in both cases. For one of these two kinds of material, the place Kaoling ("high ridge") was in ancient times in high repute; and, though it has lost its prestige since centuries, the Chinese still designate by the name " Kao-ling," the kind of earth which was formerly derived from there, but is now prepared in other places. The application of the name by Berzelius, to porcelain earth was made on the erroneous supposition, that the white earth which he received from a member of one of the embassies (I think, Lord Amherst) occurred naturally in this state. The second kind of material bears the name Pe-tun-tse ( ' ' white clay " ) . The following analyses will serve to show the average composition of (I) the natural material from King-te-Chin, such as is used in the manufacture of the finest porcelain; (II) that from the same locality used in the so-called blue Canton ware; (HI) that of the English Cor- nish or Cornwall stone; (IV) washed kaolin from St. Yrieux, France, and (V) washed kaolin from Hockessin, Delaware. 2 Constituents. I. II. III. IV. V. Silica - 73 55 73 55 73 57 48 68 48 73 Alumina Ferric oxide 21.09 18.98 16.47 27 36.92 37.02 79 Lime 2.55 1.58 1 17 16 15 1 08 21 52 11 Potash .46 I 41 Soda 2 09 | 5.84 .58 J I 04 Combined water 2.62 1.96 2.45 13.13 12.83 Total 99.62 99.70 y.i '.).s 99.83 100.09 1 American Journal of Science, 1871, p. 180. 2 Analyses I and II by J. E. Whitfield, Bulletin 27, U. S. Geological Survey; III from Langenbeck's Chemistry of Pottery; IV from Zirkel's Lehrbuch der Petrog- raphy, III, p. 758, and V by George Steiger, U. S. Geological Survey. 330 REPOET OF NATIONAL MUSEUM, Plate 15, figs. 1 and 2, will serve to show the shape and kind of the particles in the mineral kaolinite and in a prepared sample of the Hockessin kaolin, as seen under the microscope. The name halloysite is given to a white or yellowish material closely simulating kaolin in composition, but occurring in indurated masses, with a greasy feel and luster, and which adheres strongly to the tongue, a property due to its capacity for absorbing moisture. 1 As it is utilized for much the same purpose as is kaolin, it is included here. Halloysite is described by Gibson 2 as occurring in a bed some 3 feet in thickness, lying near the base of the Lower Siliceous (L. Carbon- iferous) formation, a little above or close to the Black Shale (Devonian), in Murphrees Valley, Alabama. This bed has been worked with satis- factory results near Valley Head, in Dekalb County. The present writer has found the material in comparatively small quantities, asso- ciated with kaolin, in narrow veins in the decomposing gneissic rock near Stone Mountain, Georgia. A similar occurrence is described near Elgin, Scotland. (Analysis below.) Near Tiiffer, Styria, halloy- site is described 3 as occurring in extensive thick and veinlike agglom- erations in porphyry. It is quite pure, and in the form of -irregular nodules of various sizes, frequently with a pellucid, steatitelike cen- tral nucleus, passing outwardly into a pure white substance, greasy to the touch, m which are occasionally included minute pellucid granules. Outside it passes into an earthy, friable substance. The following analyses show the varying composition of halloysite from (I) Elgin, Scotland, (II) Steinbruck, Styria, and (III) Detroit Mine, Mono Lake, California. Constituents. I. II. III. Silica 39 30 40 7 Alumina Lime 38.52 75 38.40 60 38.4 6 Magnesia. 83 Ferric oxide 1 42 Manganese 25 Water 19 34 18 00 99.20 A white chalky halloysite from the pits of the Frio Kaolin Mining Company in Edwards County, Texas (Specimen No. 53253, U.S.N.M.), 1 This property is characteristic of nearly all clay compounds when they are dry. It is to this same property that many of the so-called "madstone" owe their imagi- nary virtues. Nearly all the stones of this type examined by the writer have proved to be of indurated clay, halloysite, or a closely related compound. When applied to a fresh wound, such adhere until they become saturated with moisture, when they faU away. Their curative powers are of course wholly imaginary. 2 Geological Survey of Alabama. Report on Murphrees Valley, 1893, p. 121 "Mineralogical Magazine, II, 1878, p. 264. Report of U. S. National Museum, 1899,-Merri PLATE 15. Fig. 2. MlCROSECTIONS SHOWING THE APPEARANCE OF (0 KAOLINITE AND (2) WASHED KAOLIN. The enlargement is the same in both cases. Report of U. S. National Museum, 1899,-Merr PLATE 16. ,*v MlCROSECTIONS SHOWING THE APPEARANCE OF (D HALLOYSITE AND (2) L.EDA CLAY. The enlargement is the same in both cases. THE NONMETALLIC MINERALS. 331 has the composition given below as shown by analyses made in the laboratory of the department: Silica ._ 45.82 Alumina 39. 77 Potash 30 Ignition 13.38 99.27 The material is somewhat variable, corresponding in part to the halloysite described by Dana, and being nonplastic, and in part being plastic to an extraordinary degree. The plastic portions are almost as gritless as starch paste. Its appearance under the microscope is shown in Plate 16, fig. 1, the interspaces of the visible angular par- ticles being occupied by the past} 7 , almost amorphous material. The particles themselves act very faintly on polarized light, and it is not possible to determine their mineralogical nature. The name Indianaite has been given by Cox to a variety of halloy- site found in Lawrence County, Indiana, and which he regarded as resulting from the decomposition of Archimedes (Lower Carbonifer- ous) limestone. It is represented as forming a stratum from 6 to 10 feet thick, underlying a massive bed of Coal Measure conglomerate 100 feet thick and overlying a bed of limonite 2 to 5 feet thick. The material like kaolin is used in the manufacture of porcelain ware. The composition of this material as given by Dana is as follows: Sil- ica 39 per cent, alumina 36 per cent, water 23.50 per cent, lime and magnesia 0.63 per cent, alkalies 0.54 per cent; 99.67 per cent. (See Specimens, Nos. 29714, 34441, U.S.N.M.) The potters' and pipe clays belong mainly to what are known geologically as bedded clays, and are as a rule very siliceous com- pounds, carrying in some instances as much as 50 per cent of free quartz and 6 to 10 per cent of iron oxides and other impurities. They are highly plastic and of a white to blue, gray, or brown color (See Specimens, Nos. 17245, 33975, 20286, 67796, to 67798, from the United States and England) and burn gray, brown, or red. The tables on page 349 will show the varying composition of materials thus classed. The fire clays, so called on account of the refractory nature, differ mainly in the small percentages of lime and the alkalies they carry, and to the absence of which they owe their refractory proper- ties. (Specimens, Nos. 11629, New Jersey; 53179, Maryland; 59258, West Virginia; 68248, California; 53249-53251, South Dakoka, etc., are characteristic.) The bedded clays of the United States reach their maximum devel- opment in strata of Cretaceous and Carboniferous ages. To the Cre- taceous age belong the celebrated plastic clays of New Jersey and a very large proportion of the brick, tile, and terra cotta clays of Dela- 332 REPORT OF NATIONAL MUSEUM, 1899. ware, 1 Maryland, and Virginia. The New Jersey beds are very exten- sively utilized in Middlesex County and fully described in the State Geological Reports. 2 . As described, the entire plastic clay formation consists of several members as below, arranged in a descending series: Feet. (1) Dark-colored clay (with beds and laminee of lignite) 50 (2) Bandy clay, with sand in alternate layers 40 (3) Stoneware clay bed - 30 (4) Sand and sandy clay (with lignite near the bottom) 50 (5) South Amboy fire-clay bed 20 (6) Sandy clay (generally red or yellow) 3 (7) Sand and kaolin 10 (8) Feldspar bed 5 (9) Micaceous sand bed 20 (10) Laminated clay and sand 30 (11) Pipe clay (top white) 10 (12) Sandy clay (including leaf bed) 5 (13) Woodbridge fire-clay bed 20 (14) Fire-sand bed 15 Raritan clay beds: (15) Fire clay 15 (16) Sandy clay 4 (17) Potters' clay 20 Total 347 The following section of the Coal Measure clays at St. Louis, as pub- ilshed in Bulletin No. 3 of the Geological Survey of Missouri, will serve to show the alternating character of these beds, and their vary- ing qualities as indicated by the uses to which they are put. 3 (1) Loess, 20 feet. (2) Limestone (Coal Measure), 5 feet, (3) Clay, white and yellow, used for sewer-pipe manufacture, called "bastard fire clay," 3 to 4 feet. (4) Clay, yellow and red, sold for paint manufacture and for coloring plaster and mortar, called "ochre," 3 feet. (5) Clay, gray to white, used for paint manufacture and filling, 1 foot 6 inches. (6) Pipe clay, variegated, reddish brown and greenish, called "keel," 12 feet. (7) Sandstone. (8) Slaty shale. (9) Coal. (10) Fire clay, becoming sandy toward the base. When first mined these Coal Measure clays are usually very hard, but on exposure to the weather slack and fall into powder. They are l This of course does not include the kaolin deposits of Hockessin, Newcastle County, and similar deposits. 2 Report on Clay Deposits of Woodbridge, South Amboy, and other places in New Jersey, 1878. 3 Bulletin No. 3, Geological Survey of Missouri, 1890. Report of U. S. National Museum, 1 899. Merrill. PLATE 17. Fig. 1. % * Fig. 2. MlCROSECTIONS SHOWING THE APPEARANCE OF (1) ALBANY COUNTY, WYOMING, CLAY AND (2) FULLER'S EARTH. The enlargement is the same in both cases. Report of U. S. National Museum, 1 899.- Merrill PLATE 18. THE NONMETALLIC MINERALS. 333 as a rule much less fusible than are the glacial or stratified clays, and are used mainly in the manufacture of fire brick, sewer pipe, terra cotta stoneware, as crocks, fruit jars, jugs, etc., glass and gas retorts, smelting pots, etc. Some of these articles are made direct from the natural clays, while others are from a mixture of several clays or of a clay mixed with powdered quartz and feldspar. For ordinary brick-making purposes a great variety of materials are employed; in some cases residuary deposits, and in others alluvial and sedimentary. Throughout the glacial regions of the United States a fine unctious blue-gray material, laid down in estuaries during the Champlain epoch, the so-called Leda clays, are the main materials used for this purpose. Such are also sometimes used in making the cheaper kinds of pottery. The bowlder clays of the glacial regions are also sometimes used when sufficiently homogeneous. The prevailing colors of the Leda clays are blue -gray or yellowish. They all carry varying amounts of iron, lime, magnesia, and the alkalies, and when burned turn to red of varying tints. They fuse with comparative ease and are used mainly for brick and tile making and for the coarser forms of earthenware, such as flower pots, being as a rule mixed with siliceous sand to counteract shrinkage. The mining of such material is of the simplest kind, and consists merely of scraping away the overlying soil and sand, if such there be, and remov- ing the clay in the form of sidehill cuts or open pits. Plate 18, facing this page, shows a cut in one of the beds at Lewiston, Maine. The material here is fine and homogeneous, of a blue-gray color, and contains no appreciable grit. It is mixed with siliceous sand and used for making bricks, baking red. An analysis of the material in its air-dry state yielded results as below: Silica (SiO 2 ) 56. 17 Alumina ( A1 2 O 3 ) 24. 25 Ferrous oxide (FeO) 3. 54 Lime(CaO) 2.09 Magnesia (MgO) 2.57 Potash(K 2 O) 4.06 Soda (Na,O) 2.25 Ignition (H,O) 4.69 99.62 Under the microscope these clays are seen to be made up of beauti- fully fresh, angular bits of quartz, feldspar, mica, hornblende, and augite, with more rarely tourmalines, zircons, and other refractory minerals, with a basis of extremely fine undetermined material which may perhaps be kaolin, though the general structure of the clay is such as to suggest it owes its origin mainly to mechanical trituration, rather than chemical decomposition. The appearance of the Lewiston clay under the microscope is shown in Plate 16, fig. 1. (See Specimens 334 REPORT OF NATIONAL MUSEUM, 1899. Nos. 73036, 61041, and 61042, of these clays in their natural, mixed, and baked condition.) One of the most constant distinctions between the so-called clays of glacial and nonglacial origin, are the relatively large amounts, in the first mentioned, of lime car- bonate and alkalies and the extremely finely comminuted siliceous material to which the name rock flour is commonly given. Prof. W. O. Crosby, has shown that the smooth and plastic bluish-gray brick clays of West Cambridge contain only from one-fourth to one-third their bulk of the clay kaolin, the remainder being largely rock flour. [Proceedings of the Boston Society of Natural History, XXV, 1890.] Leda clays from Beaver County, Pennsylvania, used in the manu- facture of terra cotta at New Brighton, are reported x as having the following composition: Silica 46.160 67.780 26. 976 16.290 7.214 4.670 Titanic acid .740 .780 2.210 600 Magnesia 1.620 .727 Alkalies 3 246 2 001 Water 11.220 6.340 99.286 99.088 Vitrified brick for street pavements are made from fusible clays, sometimes in their natural condition and sometimes mixtures of ground shale and clay. (See Specimens, Nos. 61141, 61142, and 68049, from Evansville, Indiana.) The following analyses of the materials used by the Onondaga Vit- rified Pressed Brick Comjj *\y show the character of the materials there used: 2 Constituents. Calcareous layer in shale bank. A green brick; be- ing a mix- ture of the different shales. Red shale. Blue shale. Clay. Silica 25 40 57 79 45 35 Peroxide of iron. . . 2 24 6 55 5 20 4 41 Lime . Magnesia 10 39 4 67 6 38 Carbonic acid . . . Potash. Soda Water and organic matter Oxide of manganese Total OQ ^Q _ The name slip clay is given to a readily fusible, impalpably fine clay used for imparting a glaze to earthenware vessels. These clays carry 1 Second Geological Survey of Pennsylvania, Report of Chemical Analyses, p. 257. 2 Bulletin of the New York State Museum, III, No. 12, March, 1895. Clay Indus- tries of New York, p. 200. THE NONMETALLIC MINERALS. 335 iron oxides, potash and soda, together with lime and magnesia in such proportions that they vitrify readily, forming thus an impervious glass over those portions of the ware to which they are applied. The following analyses show (I) the composition of a slip clay used in pottery works in Akron, Ohio, and (II) one from Albany, New York. (Specimen No. 53583, U.S.N.M.): Constituents. (I.) (II.) Silica '. 60.40 58.54 10 42 15 41 5.36 3.19 Lime 9 88 6 30 Magnesia Alkalies 4.28 0.87 3.40 4.45 Sulphuric acid Phosphoric acid 0.65 0.09 1.10 Carbonic acid and water 8.05 8.08 Total 100.00 100.47 The Albany clay is stated by Nason * to glaze at comparatively low temperatures and to rarely crack or check. It occurs in a stratum 4 to 5 feet thick. It is used very extensively in the United States, and has even been,, shipped to Germany and France. (See also Specimens Nos. 53582, U.S.N.M., from Brimfield, Ohio; 53580, U.S.N.M., from Rowley, Michigan, and 52985, 52995, U.S.N.M., from Meissen, Saxony.) The name adobe is given to a calcareous clay of a gray-brown or yellowish color, very tine grained and porous, which is sufficiently friable to crumble readily in the lingers, and yet has sufficient coher- ency to stand for many years in the form of vertical escarpments, without forming appreciable talus slopes. It is in common use through- out Arizona, New Mexico, and Mexico proper for building material, the dry adobe being first mixed with water, pressed in rough rectangular wooden molds some 10 by 18 or more inches and 3 or 4 inches deep, and then dried in the sun. In some cases chopped straw is mixed with it to increase its tenacity. Buildings formed of this material endure for generations and even centuries in these arid climates. The material of the adobe is derived from the waste of the surrounding mountain slopes, the disintegration being mainly mechanical. According to Prof. I. C. Russell it is assorted and spread out over the valley bottoms by ephem- eral streams. It consists of a great variety of minerals, among which quartz is conspicuous. The chemical nature of the adobes vary widely, as would naturally be expected, and as is shown in the following analyses from Professor Russell's paper: 2 1 Forty-seventh Annual Report of the State Geologist of New York, 1893, p. 468. 2 Subaerial Deposits of North America, Geological Magazine, VI, 1889, pp. 289 and 342. 336 BEPOBT OF NATIONAL MUSEUM, 1899. Analyses of adobe. Constituents. I, Santa Fe, New Mexico. II, Fort Win- gate, New Mexico. III, Humboldt, Nevada. IV, Salt Lake City, Utah. SiO 2 66.69 26. 67 44.64 19. 24 A1 2 3 14.16 0.91 13.19 3.26 FeoOn 4 38 0.64 5.12 1.09 MnO 0.09 Trace. 0.13 Trace. 2 49 36.40 13.91 38.94 MgO - - - 1.28 0.51 2.% 2.75 KoO 1 21 Trace. 1.71 Trace. NaO 0.67 Trace. 0.59 , Trace. co 0.77 25.84 8.55 29.57 p 2 O 5 0.29 0.75 0.94 0.23 SO 3 0.41 0.82 0.64 0.53 Cl 0.34 0.07 0.14 0.11 H 2 O 4.94 2.26 3.84 1.67 2.00 5.10 3.43 2.96 Total 99.72 99.97 99.79 100.35 The name loess is given to certain quaternary surface deposits closely simulating adobe, but concerning the origin of which there is considerable dispute. Deposits in the United States are, according to the best authorities, of subaqueous origin. Clays of this nature are, as a rule, higher in silica than the adobes and correspondingly poorer in alumina. Loess is a common surface deposit throughout the Missis- sippi Valley, and is in many instances of such consistency as to be utilized for brickmaking. The analyses given below are from Professor Russell's paper: Analyses of the loess of the Mississippi Valley. Constituents. No.l. No. 2. No. 3. No. 4. SiO 2 72 68* 64 61 74 46 60 69 A1 2 O 3 12 03 Fe0, FeO 3.63 96 2.61 3.25 2.61 TiOjj PaO 6 . . . 23 MnO CaO 1 59 MgO 1 11 3 69 1 12 4 56 NaoO 1 68 K0 H 2 a2 50 co so, c ' ' U. lo Total... _. a Contains H of organic matter, dried at 100 C. THE NONMETALLIC MINERALS. 337 The name fullers' earth (Walkerde, volaorde, terre a foulon, terra da purgatori, etc.) includes a variety of clay of a greenish white, greenish gray, olive and oil green or brownish color, very soft, with a greasy feeling. It falls into powder in water, imparts a milky hue to the liquid, and appears to melt on the tongue like butter. It was for- merly used by fullers to take the grease out of cloth; hence the name. The English beds, according to Geikie 1 occur in Jurassic and Cre- taceous formations. Fullers' earth from beds at Nutfield, near Red- hill, Surrey, England, is described 2 as a heavy blue or yellow clay, with a greasy feel and an earthy fracture. When examined with a microscope it is found to consist of extremely irregular corroded particles of a siliceous mineral which in its least altered state is colorless, but which in nearly every case has undergone a chloritic or talcose alteration whereby the particles are inverted into a faintly yellowish green product almost wholly on polar- ized light. The particles are of all sizes up to 0.07 mm. The larger portion of the material is made up of particles fairly uniform in size and about the dimensions mentioned. In addition to these are minute colorless fragments down to sizes 0.01 mm. and even smaller. The minute size of these colorless particles renders a determination of their mineral nature practically impossible. But the outline of the cleavage flakes is evidently suggestive of a soda lime feldspar. The high percentage of silica in the insoluble residue would indicate the presence of a considerable amount of free quartz. This, however, the microscope only partially substantiates, very few of the particles showing the brilliant polarization colors characteristic of this mineral. When the powder is treated with hydrofluorsilicic acid it yields abundant crystals of potassium and aluminum fluosilicate, together with radiating forms of calcium fluosilicate. The material differs from that last described in that its particles are much larger and more angular in outline and the various elements in a different state of com- bination. (See Plate 17, fig. 2.) A substance recently put upon the American market as a fullers' earth (Specimen No. 62737, U.S.N.M., from Enid, Oklahoma), under the trade name of "glacialite," has the following chemical composition, the material being dried at 100 C. before analyzing: Silica 50. 36 Alumina 33. 38 Ferric oxide 3. 31 Sodium, lithium, potassium oxide 88 Water 12. 05 Organic matter Trace. Titanium . . . . Trace. 99.98 *Text book of Geology. 3d. ed. p. 133. 2 Geological Magazine, VI, 1889, p. 4- KC 'NAT MUS 99 22 338 KEPOKT OF NATIONAL MUSEUM, 1899. This material when placed in water falls away to a loose flocculent pow- der, which shows up under the microscope in the form of sharply angu- lar colorless particles, very faintly doubly refracting, without crystal outlines or other physical properties, such as will determine their exact mineral nature. The particles are of all sizes, from the larger floccu- lent masses, some 0.25 mm. in greatest diameter, down to those too small for measurement. The greater number lie between 0.005 and 0.01 mm., though a very large proportion are even smaller, not exceeding 0.002 mm. These smaller particles are angular in outline and almost perfectly colorless. Their appearance under the microscope is some- what that of decomposed cherts. In addition to the faintly doubly refracting particles above men- tioned, there are occasional clear, colorless, sharply angular particles of a doubly refracting mineral which can only be referred to quartz. A few yellowish iron-stained particles are suggestive of residual prod- ucts from decomposition of iron magnesian silicates. The Gadsden County, Florida, fullers' earth (Specimens Nos. 53254 and 53255, U.S.N.M.) is a light-gray material, often blackened by organic matter, and which shows under the microscope the same greenish, faintly doubly refracting particles as does the English, inter- mixed with numerous angular particles of quartz. This earth is quite plastic and sticky when wet. A section of the beds at the pits of the Cheesebrough Manufacturing Company, as given in The Mineral Resources for 1895-96, is as follows: Soil inches. . 18 Red clay feet 3 Blue clay do 3 Fullers' earth do 5J Sandy blue earth do 3 Fullers' earth (second bed) Report of U. S. National Museum, 1 899. Merrill. PLATE 19. THE NONMETALLIC MINERALS. 339 The following table l as compiled by Dr. Ries shows the variable char- acter of the material from different sources: j, 4 it J>^ S^ Si on AS-1 S Constituents. e from Cilly. a earth from R gate, b e from Steindc fel.c earth from En land, d earth from En land.e earth from Ga aunty, Florida earth from Dec unty, Georgia. OQ 1 *IU rs' earth fro east of Riv ion, Florida, h earth from I i Mount Pleasa orway, Floridi earth from ne vay, Florida../ 1 1 1 m I 1 1 11 Fullers tur Co ill P-H |!i |l SiO 2 51 21 53.00 50.17 44 00 44.00 62.83 67.46 58 72 50.70 58.30 54.60 A1.O 3 12.25 10.00 10.66 11.00 23.06 10.35 10.08 16.90 21.07 10.63 10.99 FejO 3 2.07 9.75 3.15 10.00 2.00 2.45 2.49 4.00 6.88 6.72 6.61 CaO 2.13 .50 .25 5.00 4.08 2.43 3.14 4.06 4.40 1.71 6.00 MgO 4 89 1 25 2 00 2 00 3 12 4 09 2 56 30 3 15 3 00 H 2 27.89 24.00 35.83 24.95 7.72 5.61 8.10 9.60 9.05 10.30 Na<>O 5.00 0.20 1 KO 74 j. 2.11 Moisture 6.41 6.28 2.30 7.90 9.55 7.45 Total 100.44 98.50 100.06 77.00 100.09 96.25 99.15 98.75 100.85 99.11 98.95 grE.J. Riederer, analyst. ^Standard Oil Company's property, E. J. Rie- derer, analyst. iHowell property, E. J. Riederer, analyst, j Morgan property, E. J. Riederer, analyst. aPogg. Ann., LXXVII, 1849, p. 591. b Klaproth. Beitr., Vol. IV, 1807, p. 338. c Dana, System of Min., 1893, p. 695. dGeikie,1893,p.!33. e Penny Encyclopedia, XI, Dr. Thompson, analyst. /P. Fireman, analyst. Properties of clay. To what the peculiar properties displayed by the clays are due can not as yet be said to have been fully determined. This is particularly the case with the property of plasticity and that of becoming indurated when dried. "Various explanations have been offered, but none are yet advanced which make clear all points. It has been ascribed to the impurities, to the alumina, to the combined water, and to other causes, against each of which, examples can be cited that seem to set it aside as inadequate. The impurities do not appear to cause the plasticit}^, for the sand acts unfavorably to it. The alumina is not responsible, or kaolins would be the most plastic of all, while the flint clays of Ohio are many of them approximately pure kaolins, and at the same time eminently non-plastic. 2 The combined water exerts some influence it is evident, as its expulsion entails permanent loss of plasticity, but it can not be the sole cause of plasticity, as clays equally hydrated are just as liable to differ in this respect as to agree. No theory is so well received at present as that advanced by Cook. He shows that the microscope reveals a crystalline structure which the eye does not detect, and that this structure varies greatly in degree of perfection in different samples. Some are composed of masses of 1 Seventeenth Annual Report of the U. S. Geological Survey, 1895-96, p. 880. * As is also kaolinite, the theoretically pure hydrous silicate of aluminum corre- sponding to the formula Al ? O ? .2SiO ? .2H 3 O, 340 REPORT OF NATIONAL MUSEUM, 1899. hexagonal plates or scales piled up in long bundles or faces and masses of unattached scales nearly perfect. Such clays are always but little plastic, but may become so on mechanical treatment such as grinding and kneading; on re-examination the clay then shows the same ele- ments of structure, but broken and confused, no bundles left intact, scales broken and a homogenous matrix of the crushed material derived from the still crystalline part. Clays are found in all states of this breaking up, from the highly crystalline mass to the homogenous matrix showing no plates at all; and on the degree in which the crys- talline structure is retained, its plasticity depends. This theory is cer- tainly plausible, and is supported by the fact that we always subject our clays to secure increased plasticity to mechanical disturbance which has the effect that the microscope reveals. This view harmon- izes with more points than any other advanced as yet, and offers a fair solution of the different degrees of plasticity which plastic clays exhibit, but it does not explain, nor attempt to explain, the differences which exist between flint clays and plastic clays, as Professor Cook's exami- nations were entirely confined to the latter. 1 According to Russian authorities quoted by Ries, 2 the plasticity is not only due to the interlocking of the clay particles, but varies with the fineness of the grain, the extremely coarse and fine varieties having less plasticity than those of intermediate texture. This view is also held by Drs. Ries and Wheeler. So far as the compiler's own observations go, plasticity is not de- pendent wholly upon hydration nor size nor shape of the constituent particles. The glacial (Leda) clays are made up of fresh, sharply an- gular particles of various minerals and contain less than 5 per cent com- bined water; yet in their natural condition they are extremely plastic, and scarcely less so when mixed with two-fifths their bulk of ordinary siliceous sand, as is done in the process of brickmaking. The Albany County, Wyoming, clay (Specimen No. 53229, U.S.N.M.), on the other hand, equally or even more plastic and exceedingly pasty, is made up of extremely minute particles of fairly uniform size, scarcely angular, and apparently all of the same mineral nature throughout. This yields some 16 per cent of water, on ignition, as shown in analysis, p. 348. On the whole, the evidence seems to show that the plasticity is due to the manner in which the particles conduct themselves toward moisture, and this is apparently dependent upon the size and shape and the propor- tional admixture of varying sizes of the constituents rather than upon their chemical composition. The work now being done by Dr. Whit- ney, of the Agricultural Department, on the relationship of soils to moisture bids fair to throw important light upon this branch of the subject. 1 Geological Survey of Ohio, Economic Geology, V, pp. 651-652. 2 Clay Deposits and Clay Industry in North Carolina, Bulletin No. 13, North Caro- lina Geological Survey, 1897. THE NONMETALLIC MINERALS. 341 The expulsion of the combined water in a clay is. nearly always accompanied by a diminution in volume, which varies directly as the water, or the purity of the clay. Pure kaolin shrinks as much as one- fourth of its bulk, it is stated, sometimes even more. The sandy clays used in making sewer-pipe and stoneware shrink from the tempered state from one-ninth to one-sixteenth, usually about one-twelfth. The shrinkage of the raw clay would be very much less, probably not over 3 or 4 per cent. A clay, when all the water of crystallization is expelled, will not shrink any more at red heat, but with increased heat will shrink more and more up to the moment of fusion. A pure kaolin apparently shrinks when heated a second time, even if the water is all expelled by the first heat, yet it is practically impossible to fuse it. But a good flint clay containing some sand will lose all shrinkage on being once calcined at white heat. Such clay is then used to counteract shrinkage in a body of green clay, as this effect is obtained by mixing in sand or some nonshrinking body. Many clays contain sand enough naturally to shrink little or none on heating, and some are so sandy as to actuall} T expand, though usually at the expense of soundness of struc- ture; for the particles of clay will shrink away from the grains of sand and this renders the structure very friable. The qualifications of a clay for common pottery and building mate- rial are simple, viz., plasticity when wet, and solidity and hardness when burned, but those products involving the highest qualities of clay, refractoriness, require much sharper tests. The first requisite is purity, at least purity within limits, and though the other points, density, plasticity, and non-shrinkage add greatly to the value of a pure clay, they can in no degree supply its place. Infusibility in clays rests in the aluminous base and the quartz. Long and intense heat applied to an intimate mixture of clay and silica is apt to result in a silicate of another ratio of base to acid, and which is likely to be fusible. But the great trouble with free silica in clay, in a fine state of division, is the fact that any fluxing agent read- ily unites with it, and makes a fluid slag; and in a refractory body the fusing of any one part is the beginning of the end. The constituents tending to make a clay fusible are iron, the alkalies soda and potash, and lime and magnesia. It is hard to state which is of the most consequence. Of the first two, iron is not so powerful a flux as potash, which is the worst of all the common elements; but the iron is present in larger amounts than potash in most clays, and consequently does as much harm, if not more. The effect of the iron is detrimental to the appearance of clay ware, and consequently has a direct bearing on the price of goods, while potash shows no more on the surface than on the inside, and when 342 KEPOET OF NATIONAL MUSEUM, 1899. present in the. usual small amounts it produces an incipient vitrifica- tion which makes the ware ring like a bell when struck, and is often a help in selling. The extent to which iron may be present without detriment is a point on which authorities do not agree. The Stourbridge clay of England, acknowledged to be the most refractory clay known, has 2.25 per cent of iron on an average of 100 analyses, with extremes of 1.43 and 3.63. Gros Almerode clay, has 2.12, Coblentz, 2.03, New Castle, 2.32, and yet all these clays are famous. Test mixtures of iron and pure kaolin have been run higher than this and have stood well, but as a general rule it is unsafe to rely for fine qualities on a clay with over 2 per cent of iron, particularly if the other impurities are developed in any amount. It is a well-known principle in chemistry that mixtures of bases are much more active fluxes than an equal amount of any one base; so with iron, its effect shows worse when in presence of other fluxing agents. The state in which the iron is present makes some difference; if as the sesquioxide, it takes more heat than when in the protoxide state to combine in the clay, for iron will only combine with silica in the protoxide state, and if that state is already developed, it is easier to combine the sand and iron than if in the other oxide. Sulphide of iron has a bad effect on the clay since its decomposition gives rise to the lower oxide of iron, besides the effect which the sul- phur may have. Silicate of iron is also detrimental, since it melts at a comparatively low temperature. On a piece of ware, iron in the uncombined state imparts a buff or red color; when combination begins and progresses the ware is of a bluish-gray cast, deepening as the fusion of the iron proceeds, and running to glassy black if much iron is present. Lime and magnesia act as fluxes on clays, but in any but the glacial clays the comparatively small amounts present makes them but little thought of as detrimental. They are probably present as silicates, and as these are readily fusible, their action is evidently unfavorable. When these bases are present as carbonates they combine at a higher temperature than iron or potash. The Milwaukee bricks, as already noted, are full of carbonates of lime and magnesia, and require a very hot burn, but when once the lime and silica combine they destroy the effect of 5 per cent of iron, enough to make the clay perfectly black. A brick of this kind presents an even, fine-grained, vitrified appear- ance on its fracture. 1 'They (lime and magnesia) have also the remarkable property of uniting with the iron ingredient to form a light-colored alumina-lime-magnesia-iron silicate, and thus the product is cream-colored instead of red. Mr. Sweet has shown by analysis that the Milwaukee light-colored brick contain even more iron than the Madison red brick. At numerous points in the Lake region and in the Fox River valley cream- THE NONMETALLIC MINERALS. 343 The amount of potash which a clay can contain and keep its fire properties is variously put by different authorities. As with iron, pure kaolin will stand a good deal when no other base is present, but a multi- plicity of bases makes fusion easy. Titanic acid is regarded as neutral to fire qualities; the form in which it is present being infusible. Testing clays. The statement of the tendencies and comparative power of the dangerous impurities of clay would lead us to believe we could use predictions as to their result in a given clay with some con- fidence, but the best practice does not yet trust to analysis alone. The most complete test of a clay now known would be obtained by use of such analysis as has been described, coupled with a fire test made especially to develop such points as the analysis indicates to be weak ones. Fire tests are of two kinds one is subjecting the clay to absolute heat without the action of any accompaniments, and the other is in putting the clay through the course of treatment for which it is designed to be used. The former develops the absolute quality of the clay as good or bad, the latter proves or disproves the fitness of the clay for the work. The latter is better of course as a business test wherever it is practicable to use it. The former can be made only in a specially adapted furnace. The clay is cut into one-inch cubes with square edges, and is set in a covered crucible resting on a lump of clay of its own kind, so that it touches no foreign object. The heat is then applied, and its effect will vary from fusing the mass to a button to leaving it with edges sharp and not even glazed on the surface. Expe- rience soon renders one proficient in judging of clays by this test. 1 A method of testing the fusibility of clays by comparing them with samples of known composition and fusibility has of late years come into extensive use. These prepared samples, known from their inventor and their shape as Seger's pyramids, consist of mixtures in varying proportions of kaolin and certain fluxes, so prepared that there is a constant difference between their fusing points. When such pyramids, together with the samples to be tested, are placed in a furnace or kiln, colored brick are made from red clays. In nearly or quite all cases, whatever the original color of the clay, the brick are reddish when partially burned. The expla- nation seems to be that at a comparatively moderate temperature the iron constit- uent is deprived of its water and fully oxidized, and is therefore red, while it ia only at a relatively high heat that the union with the lime and magnesia takes place, giving rise to the light color. The calcareous and magnesian clays are, therefore, a valuable substitute for true aluminous clays, for they not only bind the mass together more firmly, but give a color which is very generally admired. They have also this practical advantage, that the effects of inadequate burning are made evident in the imperfect development of the cream color, and hence a more carefully burned pro- duct is usually secured. It is possible to make a light-colored brick from a clay which usually burns red by adding lime. The amount of lime and magnesia in the Milwaukee brick is about 25 per cent. In the original clays in the form of carbo- nates they make up about 40 per cent. Geology of Wisconsin, I, 1873-79, p. 669.) 1 Geological Survey of Ohio, Economic Geology, V, pp. 652-655. 344 REPORT OF NATIONAL MUSEUM, 1899. they begin to soften as the temperature is raised, and as it approaches their fusion point the cones bend over until the tip is as low as the base. When this occurs the temperature at which they fuse is con- sidered to be reached. 1 Uses. Clay when moistened with water is plastic and sufficiently firm to be fashioned into any form desired. It can be shaped by the hands alone; by the hands applied to the clay as it turns with the pot- ter's wheel, or it can be shaped by moulds, presses or tools. When shaped and dried, and then burned in an oven or kiln, it becomes firm and solid, like stone; water will not soften it, it has entirely lost its plastic property, and is permanently fixed in its new forms, and for its designed uses. These singular and interesting properties are possessed by clay alone, and it is to these it owes its chief uses. It is used (1) for making pottery; (2) for making refractory materials; (3) for making building materials; (4) for miscellaneous purposes. Pottery. Pure clay worked into shapes and burned, constitutes earthenware. The ware of itself is porous, and will allow water and soluble substances to soak through it. To make it hold liquids, the shaped clay before burning is covered with some substance that in the burning of the ware will melt and form a glass coating or glazing which will protect the ware in its after uses from absorbing liquids, and give it a clean smooth surface. The color of the ware depends on the purity of the clay. Clays containing oxide of iron burn red, the depth of color depending on the amount of the oxide, even a small fraction of 1 per cent being sufficient to give the clay a buff color. Clay containing oxide of iron in sufficient quantity to make it par- tially fusible in the heat required to burn it, when made into forms and burned, is called stoneware clay. The heat is carried far enough to fuse the particles together so that the ware is solid and will not allow water to soak through it; and the fusion has not been carried so far as to alter the shapes of the articles burned. The oxide of iron by the fusion has been combined with the clay, and instead of its characteristic red, has given to the ware a bluish or grayish color. Stoneware may be glazed like earthenware, or by putting salt in the kiln, when its vapor comes in contact with the heated ware and makes with it a sufficient glaze. Clay which is pure white in color and entirely free from oxide of iron, may be intimately mixed with ground feldspar or other minerals which contain potash enough to make them fusible, and the mixture still be plastic so as to be worked into forms for ware. When burned, such a composition retains its pure white color, while it undergoes 'See Dr. Ries's paper on North Carolina clays, already quoted, and also his numerous contributions on their subject in the volumes of the United States Geolog- ical Survey relating to mineral statistics. THE NONMETALLIC MINERALS. 345 fusion sufficient to make a body that will not absorb water. And its surface can be made smooth and clean by a suitable plain or orna- mented glaze. Ware of this kind is porcelain or china. The analyses on page 349, compiled from works believed to be authori- tative, show the varying character, so far as chemical composition is concerned, of the clays. In most of the analyses, it will be observed, the silica existing in the form of quartz is given in a separate column from the combined, while in column 4 is given the actual calculated percentage of kaolin which the analyses indicates each sample contains. Refractory materials. Modern improvements in metallurgy, and in furnaces for all purposes, are dependent to a great degree on having materials for construction which will stand intense heat without fusing, cracking, or yielding in any way. The two materials to which resort is had in almost all cases, are pure clay, and quartz in the form of sand or rock. They are both infusible at the highest furnace heats. The clay, however, is liable to have in it small quan- tities of impurities which are fusible, and it shrinks very much when heated to a high temperature. Quartz rocks are very liable to crack to pieces if heated too rapidly, and both the rocks and sand are rap- idly melted when in contact with alkalies, earths or metallic oxides, at a high temperature. They do not shrink in heating. Sandstone, or quartz rock, is not as much used as a refractory material as it was formerly. Bricks to resist intense heat are made of cla} T , of sand, and of a mixture of clay and sand. The different kinds are specially adapted to different uses. Fire bricks made of clay, or clay and sand, are the ones which have been generally made in the United States. To make these, the clay which stands an intense heat the best, is selected as the plastic mate- rial of the brick. This is tempered so that it may not shrink too much or unevenly in burning, by adding to the raw clay a portion of clay which has been burned till it has ceased to shrink and then ground, or a portion of coarse sand, or a quantity of so-called feld- spar. These materials are added in the proportions which the experi- ence of the manufacturer has found best. The formula for the mixture is the special property of each manufacturer, and is not made public. The materials, being mixed together and properly wet, are molded in the same way as common bricks arei, and after they have dried a little, they are put into a metallic mould and subjected to powerful pressure. They are then taken out, dried, and burned in a kiln at an intense heat. It does not appear which is the best for tempering, burned and ground clay, or coarse sand, or feldspar. Reputable manufacturers are found who use each of these materials, and make brick that stand fire well. It is of the utmost importance to select the materials care- 346 REPORT OF NATIONAL MUSEUM, 1899. fully, and to allow no impurity to get in while handling the clay or working the components together. Fire bricks intended, in addition to their refractory qualities, to retain their size and form under intense heat without shrinkage, have been made to some extent. The English Dinas bricks are of this kind, and the German and French "silica bricks." The Dinas bricks are of quartz sand or crushed rock, and contain very little alumina and about one per cent of lime. They stand fire remarkably well, the lime being just enough to make the grains of sand stick together when the bricks are intensely heated. In the other "silica bricks," lire clay to the amount of 5 or 10 per cent is mixed with the sand, and this plastic material makes the particles of the sand cohere sufficiently to allow of handling the bricks before burning. They have met the expectation of those who made them, and are extensively used. 1 Under the head of "Miscellaneous uses of clay," p. 317, Cook gives the following, which may well be incorporated entire: Paper clay. Clay which is pure white and that also which is discolored and has been washed to bring it to a uniform shade of color, is used by the manufacturers of paper hangings, to give the smooth satin surface to the finished paper. It is used by mixing it up with a thin size, applying it to the surface of the pieces of paper, and then polishing by means of brushes driven by machinery. The finest and most uniformly colored clays only are applicable to this use, and they are selected with great care. Clay is also used to some extent by paper manufacturers, to give body and weight to paper. Heavy wrapping paper, such as is used by the United States Post- Office Department, must, according to specifications, contain 95 per cent of jute butts and 5 per cent of clay. The cheaper forms of con- fectionery, particularly such as is sold from carts upon the streets, is very heavily adulterated with this material. Alum clay. A large quantity of clay is sold every year to the manufacturers of chemicals, for making alum. A rich clay is needed for this purpose, but those con- taining lignite or pyrite which renders them inapplicable for refractory materials, do not spoil them for this use. Alum is made by digesting the clay in sulphuric acid, which forms sulphate of alumina, then dissolving out the latter salt from the silica and other impurities, and forming it into alum by the addition of the necessary salt of potash, soda, or ammonia, and crystallizing out the alum. The white clay of Gay Head and Chilmark, Marthas Vineyard, Massachusetts, was at one time used extensively for alum making, according to Edward Hitchcock. 2 As a substitute for sand in making mortar and concrete clay is per- haps the best material to be found. For this purpose the clay is burnt so that it is produced in small irregular pieces that are very hard and durable. These pieces are then ground to a fairly fine powder, which is used to mix with the lime or cement just as sand would be. The Geological Survey of New Jersey, Report on Clay Deposits, pp. 307-312. 2 American Journal of Science, XXII, 1832, p. 37. THE NONMETALLIC MINERALS. 347 result is a very strong 1 mortar, in some cases stronger than when sand is employed. 1 The so-called gumbo clays, sticky, tough, and dark-colored clays of the Chariton River region, Missouri, are hard burned and used for railroad ballast and macadam. Under the names of -Rock Soap and Mineral Soap there have from time to time been described varieties of clay which, owing to their soapy feeling, are suggestive of soap, and which in a few instances have been actually used in the preparation of this material. A rock soap from Ventura County, California, has been described by Prof. G. H. Koenig as a mixture of sandy and clayey or soapy material in the proportion of 45 per cent of the first and 55 per cent of the second. The chemical composition of the material and of the two portions is given below: Constituents. Crude material. Sandy portion. Soapy portion. Silica 67 55 69 40 73 10 Alumina and iron Lime 12.97 77 13.50 30 14.10 Magnesia Potash 0.85 1 43 Trace. Not de- ter - Soda Water 3.63 13 67 } 4.55 12 25 mined. 6 70 Nearly all the silica is reported as being in a soluble or opalescent state and the alumina as either a hydrate or very basic silicate. It is said 2 that at one time the material was made into a variety of use- ful articles, as "salt water soap," scrubbing and toilet soap, tooth pow- der, etc. A somewhat similar material from Elk County, Nevada, has been used for like purposes, and put upon the market under the name of San-too-gah-choi mineral soap. This clay is of a drab color, with a slight pinkish tint, a pronounced soapy feeling and slight alkaline reaction when moistened and placed upon test paper. An analysis by Packard in the laboratory of the U. S. National Museum yielded: Silica 48.80 Alumina 18.57 Iron oxides 3. 88 Lime 1. 07 Magnesia 2. 52 Soda 2.32 Potash 1.12 Ignition 21.13 Total. 1 The Worlds Progress, February, 1893. 2 Sixth Annual Report of the State Mineralogist of California, 1886, Pt. 1, p. 132. 348 REPORT OF NATIONAL MUSEUM, 1899. Mention may be made here also of the material sold in the shops under the name of Bon Ami and used for cleansing glass and other like substances. This under the microscope shows abundant minute sharply angular particles, consisting of partially decomposed feldspar mixed with a completely amorphous mineral which may be opalescent silica or possibly a very finely comminuted pumice. An analysis by Packard in the laboratory of the Department yielded: Silica 59.86 Alumina 18. 74 Magnesia 0. 34 Potash 10. 70 Soda 3.51 Ignition 7. 67 Total 100.82 Alcohol extracts 7.43 per cent, and water 0.2M per cent in addition, the extract having a soapy appearance and the odor of some essential oil. A soapy clay occurring near Rock Creek station, in Alban} r County, Wyoming, has been shipped in considerable quantities during the past few years to New York, Philadelphia, and Chicago, but the use to which it was put remains a secret. It is stated l that at first the mate- rial was sold at the rate of $25 a ton, but that the price has now dropped to $5 a ton. Analyses are given as below. The chief physical characteristic of this clay, aside from its soapy feeling, is its enormous absorptive power, the -absorption being attended naturally with an increase in bulk amounting to several times that of the origi- nal mass. 2 Plate 17, fig. 11 shows the extreme fineness and homoge- neity of this clay as seen under the microscope. Constituents. I. Rock Creek. II. Crook County. HI. Weston County. rv. Natrona County. SiO 2 59 78 61 08 63 25 65 24 A1 2 O 3 15 10 17 12 12 62 15 88 Fea0 3 ... - 2 40 3 17 3 70 3 12 MgO CaO 73 2 G9 4 12 j- 5.34 Na^O K 2 O (a)' ' SO H.,O 16 26 Specific gravity 2 132 a No estimate. Engineering and Mining Journal, LXIII, 1897, p. 600; LXVI, 1898, p. 491. 2 A small plug of this clay fitted to accurately occupy a space of 20 cubic centi- meters in the bottom of a conical measuring flask, and kept saturated with water for two days, swelled to a bulk of 160 cubic centimeters. The absorption was so com- plete that none of the water ran off when the flask was inverted, and the condition of the clay resembled that of flour or starch paste. THE NONMETALLIC MINERALS. 349 Authority. IP i 1 1 . og! S^^ ^ ^ ^ oT P. 02 >^ S > tf O tft ITTW rang S3 xi2 So- SKSSw 8 giSSSSS s s S SJ S ojnjstojAf s 3 | : : S S S 3 3 "8 8 S Xjundun sno S S S 1 : S 58 ^ S s 2 S 2 (0 5 n) B TO n (O**N) Bpog 3 R !] o o d d d .4' .4' 1.9 j j .!* (0 5 H)W>d S : : 5 1 ! ! : i S ' (0 la3 w S 2 : S S g S d d do do 5 s o (OK amiT d d d r-' d r-; d jo apixoinbsag 00 * s 3 ! S 5 S | : : 3 A" p u B B imo j, ^ 8 g 3 ? : " 3 S j j : : 2 J'OUj) ( 5 S 8 S S : S : d d d d j d j : : aj ' (S z IS jn 10 * g d 5? S | S j : : 8 aqi jo jiiao jaj * p g S 3 2 S 3 53' 8 S S S S S s s ri (0 S H) J -WAV pouiqtnoo Jg S S fc SooKS ^ gj cj * .-: i X I^ uuirantv 2S SS S355S?S s a s s 2 ?5 2 a 2 3 S 2 (WS) BO -His pautqraoo - 11 1 1 1 1 1 1 1 213 Name of company ard location. flli 1.3*1 ill"! | 5 1 ai^ig I 1 S d ! fc O 5 "2 S >> ^:s > sc52 '-iS.'^tn -og fl *3 aiOo; ^Og4>tio lict lie Iliillilfi l s i w ri I f ll f l 8 P Do Do East Liverpool, Ohio (yellow-ware clay). 350 EEPORT OF NATIONAL MUSEUM, 1899. i 5^* .2" .2 >, a (A O Q js s "o 3 I lls 5 > S ^ o o l^o 6 d 8 > O o S M S w 8 8 TO ljoj rang a 8 8 S S g 1 8 S 88 g S S 8 i a3 aan^siojn ^ 1 frl rH d d s S c^ Jo ^ S Xjundrai sno S3 ci r- S d S d 5! ^j J; g s CO & i 8 S ( S O 5 9^) UOJl r-H IN l>i CN 10 16 d in ci rH rH jo gpixombsas fS ^ *ss?su - ! S g g 00 : (*OTX) 1 s e s i S : piou iu)ix 9, s Q l^ OS SB W in CO . 10 to IN + * * ( S OTS 4 8 5 -^ oqj jo juao aaj; + i S 1 5 J 9 !i s & SS (O 5 H) * -BM paujqraoo si rH s SS 5 S 00 oJ t-: S |S s s JH CO S 5S o o5 C0 5 IV) uuiumiv ei i 8 S? 1 1 1 1 53 P -r ('OSS) T?D 8 S J2 g S S S S S ^ -His pauiqraoo " ^ S S 3 8 s " S S Name of company and location. Potters' clays Continued. H. Cutter & Sons, Woodbridge, New Jersey. Bine Ball clay, Pennsylvania Pipe clay. N. U. Walker, Walker's Station, Ohio (sewer pipe) . Bolivar clay, Island Siding, Ohio (fit for pipe). W. H. Evans, Waynesburg, Ohio (drain pipe). A. O. Jones, Columbus, Ohio (drain tile). Whitmore, Robinson & Co., Akron, Ohio (kaolite slip clay). Fire clay. C. E. Holden, Mineral Point, Ohio. Scioto Fire Brick Co., Sciotoville, Ohio. Do Wassail Fire Clay Co., Columbus, Ohio. THE NONMETALLIC MINEEALS. 351 S CO S S S 3 i S 8 8 SJ i 8 3 8 S g 8 rH rH r-* o : c4 S SJ S s S 8 S C^ CO CO KJ 6 5' I o S 8 S ^ o d S o 1 c? S g a -a =S fs S y rt ^3 g ^ o 6 OS 5 a S ?! S s 2 3 s S S g ^' S S O o o : s s o ^ - 8 r4 d rt *~* 1 ^ S3 S^S . o : g si S 58S I K S S 8 S ?2 8 08 8 K s od oS o ^! M '~ S to & !J 8 S s K g & ?1 S S 8 ?i 8 S CO ?! 53 S S S S g 00 I ii: a $ ^ o -T S si 5 Island Fire Clay Co., nea benville, Ohio. Ballou clay, Zanesville, Oh Etna Fire Brick Co., Oakhi B. Ellison, south-southw Bonhamtown, New Jerse Brick clay Milwaukee brick clay, Wia< Mount Savage, Maryland.. Newcastle, England Sayre & Fisher, front brie Sayreville, New Jersey. 352 REPORT OF NATIONAL MUSEUM, 1899. The bibliography of clays is very extensive, and but a few references are given here. The reader is referred particularly to Branner's Bibli- ography of Clays and the Ceramic Arts, 1 and to the papers of Dr. H. Ries in the reports on the Mineral Resources of the United States, published annually by the U. S. Geological Survey. S. W. JOHNSON, JOHN M. BLAKE. On Kaolinite and Pholerite. American Journal of Science, XLIII, 1867, p. 351. J. C. SMOCK. The Fire Clays and associated Plastic Clays, Kaolins, Feldspars, and Fire Sands of New Jersey. Transactions of the American Institute of Mining Engineers, VI, 1877, p. 177. GEORGE H. COOK. Report on the Clay Deposits of Woodbridge, South Amboy, and other places in New Jersey. Geological Survey of New Jersey, 1878. RICHARD C. HILLS. Kaolinite, from Red Mountain, Colorado. American Journal of Science, XXVII, 1884, p. 472. See also Bulletin No. 20, U. S. Geological Survey, 1885, p. 97. J. P. LESLEY. Some general considerations respecting the origin and distribution of the Delaware and Chester kaolin deposits. Annual Report Geological Survey of Pennsylvania, 1885, p. 571. J. H. COLLINS. On the Nature and Origin of Clays: The Composition of Kaolinite. Mineralogical Magazine, VII, December, 1887, p. 205. J. FRANCIS WILLIAMS, R. N. BRACKET. Newtonite and Rectorite two new minerals of the Kaolinite Group. American Journal of Science, XLII, 1892, p. 11. EDWARD ORTON. The Clays of Ohio, Their Origin, Composition', and Varieties. Report of the Geological Survey of Ohio, VII, 1893, pp. 45-68. EDWARD ORTON, jr. The Clay Working Industries of Ohio. Report of the Geological Survey of Ohio, VII, 1893, pp. 69-254. H. 0. HOFMAN, C. D. DEMOND. Some Experiments for Determining the Refractori- ness of Fire Clays. Transactions of the American Institute of Mining Engineers, XXIV, 1894, p. 42. W. MAYNARD HUTCHINGS. Notes on the Composition of Clays, Slates, etc., and on some Points in their Contact-Metamorphism. The Geological Magazine, I, 1894, p. 36. H. JOCHUM. The Relation between Composition and Refractory Characters in Fire Clays. Minutes of Proceedings of the Institution of Civil Engineers, CXX, 1894-95, p. 431. J. A. HOLMES. Notes on the Kaolin and Clay Deposits of North Carolina. Transactions of the American Institute of Mining Engineers, XXV, 1895, p. 929. HEINRICH RIES. Clay Industries of New York. Bulletin No. 12 of the New York State Museum, III, March, 1895, pp. 100-262. JOHN CASPER BRANNER. Bibliography of Clays and the Ceramic Arts. Bulletin No. 143, U. S. Geological Survey, 1896. W. S. BLATCHLEY. A Preliminary Report on the Clays and Clay Industries of the Coal and Coal-Bearing Counties of Indiana. The School of Mines Quarterly, XVIII, 1896, p. 65. W. MAYNARD HUTCHINGS. Clays, Shales, and Slates. The Geological Magazine, III, 1896, p. 309. CHAS. F. MABERY, OTIS T. FLOOZ. Composition of American Kaolins. Journal of the American Chemical Society, XVIII, 1896, p. 909. 1 Bulletin No. 143, U. S. Geological Survey, 1896. THE NONMETALLIC MINERALS. 353 ('HAS. F. MABERY, OTIS T.FLOOZ. Clay, Bricks, Pottery, etc. Thirteenth Report of the California State Mineralogist, 1896, p. 612. THOMAS C. HOPKINS. Clays and Clay Industries of Pennsylvania. Appendix to the Annual Report of the Pennsylvania State College for 1897. J. NELSON NEVIUS. Kaolin in Vermont. Engineering and Mining Journal, LXIV, 1897, p. 189. HEINRICH RIES. The Clays and Clay- Working Industry of Colorado. Transactions of the American Institute of Mining Engineers, XXVII, 1897, p. 336. H. A. WHEELER. Clay Deposits. Missouri Geological Survey, XI. W. W. CLENDENNIN. Clays of Louisiana. Engineering and Mining Journal, LXVI, 1898, p. 456. M. H. CRUMP. The Clays and Building Stones of Kentucky. Engineering and Mining Journal, LXVI, 1898, p. 190. W. C. KNIGHT. Bentonite. [A New Clay.] Engineering and Mining Journal, LXVI, 1898, p. 491. The Building Stones and Clays of Wyoming. Engineering and Mining Journal, LXVI, 1898, p. 546. HEINRICH RIES. Physical Tests of New York Shales. School of Mines Quarterly, XIX, 1898, p. 192. The Ultimate and the Rational Analysis of Clays and Their Relative Advantages. Transactions of the American Institute of Mining Engineers, XXVIII, 1898, p. 160. EUGENE A. SMITH. The Clay Resources of Alabama and the Industries Dependent upon Them. Engineering and Mining Journal, LXVI, 1898, p. 369. J. E. TODD. The Clay and Stone Resources of South Dakota. Engineering and Mining Journal, LXVI, 1898, p. 371. VII. NIOBATES AND TANTALATES. 1. COLUMBITE AND TANTALITE. These are columbates and tantalates of iron and manganese, colum- bite representing the nearly pure colurabate and tan tali te the nearly pure tantalate. Both are likely to carry varying quantities of iron and manganese. The analyses given below will serve to show the varying composition, No. I being coluinbite from Greenland, No. II from Haddam, Connecticut, and Nos. Ill and IV from the Black Hills of South Dakota: Constituents. I. II. III. IV. Coliimhiiim ppntnxirtp 77.97 51.53 54.09 29.78 Tantalium pentoxide 17 33 28.55 13 54 18.20 11 21 53. 28 G 11 Manganese protoxide 3.28 4.55 7.07 10. 40 With traces of tin, wolfram, lime, magnesia, etc. The mineral is of an iron black, grayish or brownish color, opaque, often with a bluish iridescence, dark red to black streak, specific gravity varying from 5.3 to 7.3 and hardness of 6. Insoluble in acids. NAT MUS 99 23 354 REPORT OF NATIONAL MUSEUM, 1899. Occurrence. The mineral occurs in granitic and feldspathic veins in the form of crystals, crystalline granules, and cleavable masses. In the United States it has been found in greater or less abundance in nearly all the States bordering along the Appalachian Mountain sys- tem (Specimen No. 63478, from Portland, Connecticut), in the Black Hills of South Dakota, and also in California and Colorado. It has also been found in Italy, Bavaria, Finland, Greenland, and western South America. Uses. The material is used only in the preparation of salts of columbium and tantalium, and as but a small quantity of these salts are used, the mineral is in but little demand, except as mineralogical specimens. 2. YTTROTANTALITE. This name is given to a mineral closely related to samarskite (see below), but carrying smaller percentages of uranium and lacking in didymium and lanthanum. It is essentially a tantalate of yttrium with small amounts of other of the rarer earths. (Specimen No. 60926, U.S.N.M.) In appearance it is distinguished from samarskite only with difficulty. Pyrochlore, fergusonite, aeschynite, euxenite, etc., are closely related compounds, the commercial uses of which have not yet been demonstrated. 3. SAMARSKITE. Composition as given below. When crystallized, in the form of rectangular prisms, but occurring more commonly massive and in flattened granules. Cleavage, imperfect; fracture conchoidal; brittle. Hardness, 5 to 6; specific gravity, 5.6 to 5.8. Luster, vitreous to resinous. Color, velvet black. (Specimen No. 62772, U.S.N.M.) Constituents. I. II. III. IV. Columbic oxide - -i {37 20 61. M Tantalic oxide / 54.81 54.96 55.13 18 60 Tungstic and stannic oxides le 31 08 Uranic oxide.. 17 03 Ferrous oxide 14 07 14 02 11 74 10 90 Manganous oxide . . Cerous oxide, etc 3 95 5 17 4 24 4 25 Yttria 11 11 14 45 Magnesia Lime 52 55 Loss by ignition Insoluble 1 25 101.21 100.40 99.12 100.36 Localities and mode of occurrence. The only localities of importance in the United States are the Wiseman Mica Mine and Grassy Creek Mine, in Mitchell (Specimen No. 62772, U.S.N.M.) and in McDowell THE NONMETALLIC MINERALS. 355 counties, North Carolina. The mineral has also been found in the form of black grains and pebbles, sometimes weighing one-fourth of an ounce, in the gold-bearing sands of Rutherford County. At the Wiseman Mine large masses, one weighing upwards of 20 pounds, were found some years ago. The analyses quoted above were made from material from this mine. 1 Uses. See under Monazite, p. 383. 4. WOLFKAMITE AND HtJ3NEKITE. Composition, a tungstate of manganese, and iron. The proportion of the iron and manganese are quite variable, the tungsten remaining nearly constant. The name hiibnerite is given to the variety contain- ing very little iron, but consisting essentially of tungsten and man- ganese. The following table shows the range in composition: Locality. WO 3 . FeO. MnO. CaO. MgO. Wolframite: Adun-Chalon 75.55 21.31 2.37 0.26 0.51 Monroe, Connecticut 75.47 9.53 14.26 Hiibnerite: Bonita, New Mexico 76.33 3.82 19.72 0.13 Trace. Nye County, Nevada 74.88 0.5C 23.87 0.14 0.08 Wolframite is dark reddish brown to black in color, with a resinous luster; has a hardness of about 5, a specific gravity of 7.55, and a pro- nounced tendency to cleave with flat, even surfaces. Its great weight, color, and cleavage tendencies are strongly marked characteristics, and the mineral once identified is as a rule easily recognized. Occurrence. The mineral is found in veins associated with tin ore (cassiterite), and also with quartz, pyrite, galena, sphalerite, etc. The chief known localities in the United States are Monroe and Trumbull, Connecticut; Blue Hill Bay, Maine; Rockbridge County, Virginia (Specimen No. 65206, U.S.N.M.); the Mammoth district, Nye and Lander counties, Nevada (Specimens Nos. 15755, 5653, U.S. N.M.); Black Hills, S. Dakota (Hubnerite) (Specimen No. 53461, U.S.N.M.); Bonita and White Oaks, Lincoln County, New Mexico; Falls County, Texas (Specimen No. 62766, U.S.N.M.); Russellville, Arizona (Specimen No. 53223, U. S. N. M.). The foreign localities are the tin regions of Bohemia, Saxony (Specimen No. 67752, U.S.N.M.), and Cornwall and Devon- shire (Specimens Nos. 67460, 67753, 67787, and 67788, U.S.N.M.), England; also Australia (Specimens Nos. 60978, 60967, U.S.N.M.)and Bolivia and Peru, South America. For uses, see under Scheelite, p. 356. ^ee the Minerals of North Carolina, Bulletin 74, U. S. Geological Survey, 1891. 356 KEPORT OF NATIONAL MUSEUM, 1899. 5. SCHEELITE. This is calcium tungstate, consisting when pure of some 80.6 per cent tungsten trioxide (WO,) and 19.4 per cent lime; usually, however, carrying from 1 to 8 per cent of molybdic oxide (MoO 3 ). The min- eral is white and translucent, and yellow and brownish in color, with a hardness of 4.5-5, gravity 6, and a tendency to cleave into octa- hedral forms. The occurrence is similar to that of wolframite, but the mineral is less common. Uses. The tungstates have been used mainly in the manufacture of tungstic acid, but the metal tungsten is coming into use as an alloy in making steel. Recently attempts have been made in France to utilize the material in porcelain glazes, but thus far without much success. There is at present no regular source of supply in America. BIBLIOGRAPHY. J. PHILLP. Tungsten Bronzes. Journal of the Society of Chemical Industry, I, 1882, p. 152. The Use of Wolfram or Tungsten. Iron Age, XXXIX, 1887, p. 33. T. A, RICKARD. Tungsten. Engineering and Mining Journal, LIII, 1892, p. 448. Wolfram Ore. Iron Age, XL, 1892, p. 229. ADOLF GURLT. On a Remarkable Deposit of Wolfram Ore in the United States. Transactions of the American Institute Mining Engineers, XXII, 1893, p. 236. See also Engineering and Mining Journal, LVI, 1893, p. 216. F. CREMER. The Place of Tungsten in the Industries. Iron Age, LVI, 1895, p. 536. HENRI MOISSAN. Researches on Tungsten. Minutes of the Proceedings of the Institution of Civil Engineers, CXXVI, 1895-96, p. 481. R. HELMHACKER. Wolfram Ore. Engineering and Mining Journal, LXII, 1896, p. 153. Prof. BODENBENDER. Wolfram in the Sierra de Cordoba, Argentine Republic. Transactions of the North of England Institute of Mining and Mechanical Engineers, XLV, Pt. 3, March, 1896, p. 59. VIII. PHOSPHATES. 1. APATITE; ROCK PHOSPHATE; GUANO; ETC. Phosphorus is one of the most widespread of the elements, and is apparently indispensable to both animal and vegetable life. In nature it occurs in various compounds, by far the more common being the phosphates of calcium and aluminum, such as are commercially used as fertilizers. These in various conditions of impurity occur under THE NONMETALLIC MINERALS. 357 several forms, some distinct and well defined, others illy defined and passing by insensible gradations into one another, but all classed under the general term of phosphates. Their origin and general physical properties are quite variable, and any attempt at classifying must be more or less arbitrary. For our present purposes it is sufficient that we treat them under the heads of mineral phosphates and rock phos- phates, as has been done by Dr. Penrose. 1 These two classes are then subdivided as below: (A t't -I ^ uor a P a tites, (I) Mineral phosphates 2 .. -I _P ( Chlor apatites. I 1 nospnonte. Amorphous nodular phosphates loose or cemented into conglomerates. Phosphatic limestones. (II) Rock phosphates , , , , 1 Guanos.. J Soluble guanos. ( Leached guanos. Bone beds. APATITE. Under the name of apatite is included a mineral composed essentially of phosphate of lime, though nearly always carrying small amounts of fluorine or chlorine, thereby giving rise to the varieties fluar-a,patite and chlor-apatite. The mineral crystallizes in the hex- agonal system, forming well-defined six-sided elongated prisms of a green, yellow, rose, or reddish color, or sometimes quite colorless. (Specimens, Nos. 62128, 62129, U.S.N.M., from Renfrew, Canada.) It also occurs as a cn T stalline granular rock mass. (Specimens, Nos. 62137, 62148, 65111^ U.S.N.M.) The hardness is 4.5 to 5; specific gravity, 3.23; luster, vitreous. Apatite in the form of minute crystals is an almost universal constituent of eruptive rocks of all kinds and all ages. It is also found in sedimentary and metamorphic rocks as a constituent of veins of various kinds, and is a common accompani- ment of beds of magnetic iron ore. It is only when occurring segre- gated in veins and pockets, either in distinct crystals or as massive 1 Bulletin No. 46 of the U. S. Geological Survey. 2 Fuchs (Notes Sur la Constitution des Gites Phosphate de Chaux) divides the natural phosphates into three classes. In the first the phosphatic material is concen- trated in sedimentary beds; in the second it is disseminated throughout eruptive rocks, and in the third it constitutes entirely or partially the material filling veins and pockets. That found in sedimentary beds occurs in rounded and concretionary masses called nodules. In eruptive and metamorphic rocks the phosphate occurs in the crystalline form of apatite, sometimes isolated or grouped in aggregates. In veins the phosphate occurs massive and in pockets, crystalline, but not in distinct crystals; rather as globular and radiating masses. To such the name phosphorite is given. The three varieties show a like variation in solubility, the amorphous phos- phates being soluble in citrate or oxalate of ammonia to the extent of 30 to 50 per cent; the phosphorites to the extent of only 15 to 30 per cent, and the apatite scarcely at all. The amorphous phosphates alone have proven of value for direct application to soils, the other varieties needing previous treatment to render them soluble. 358 REPORT OF NATIONAL MUSEUM, 1899. crystalline aggregates, as in Canada and Norway, that the material has any great economic value. The average composition of the apa- tites, as given in the latest edition of Dana's Mineralogy, is as follows: Variety. P 2 5 CaO F. Cl. atite 41.0 53.8 6.8 or Ca 3 P 2 O 8 89.4 + CaCl, 10. 6. ft 42.3 55.5 3.8 orCa 3 P 2 8 92.25 + CaF 2 7.75. The name phosphorite covers a material of the same composition as apatite, but occurring in massive concretionary and mammilary forms. (Specimens No. 37147, U.S.N.M., from Spain and 66741, U.S.N.M., from Florida). The name was first used by Kir wan in describing the phosphates of Estremadura, Spain, which occur in veins and pockety masses in Silurian schists, as noted later. ROCK PHOSPHATE. The general name of rock phosphate is given to deposits having no definite composition but consisting of amorphous mixtures of phosphatic and other mineral matter in indefinite propor- tions. Here would be included the amorphous nodular phosphates like those of our Southern Atlantic States (Specimens Nos. 34322, 44244, 66737, U.S.N.M.), phosphatic limestones and marls (Specimens Nos. 62718, U.S.N.M., Africa, and 62723 Utah), guano (Speci- men No. 69281, TJ.S.N.M.), and bone bed deposits (Specimens Nos. 66581, 67332, U.S.N.M.). These are so variable in character that no satisfactory description of them as a whole can be given. The name coprolite is given, to a nodular phosphate such as occurs among the Carboniferous beds of the Firth of Forth in Scotland, and which is regarded as the fossilized excrement of vertebrate animals. (Specimen No. 62731, U.S.N.M.) Phosphatic limestones and marl, as the names denote, are simply ordinary limestones and marls containing an appre- ciable amount of lime in the form of phosphate. Such are rarely suf- ficiently rich to be of value except in the immediate vicinity, owing to cost of transportation. Guano is the name given to the accumulations of sea-fowl excretions, such as occur in quantities only in rainless regions, as the western coast of South America. The most noted deposits are on small islands off the coast of Peru. The material is of a white-gray and yellowish color, friable, and contains some 20 or more per cent of phosphate of lime, 10 to 12 per cent of organic matter, 30 per cent ammonia salts, and 20 per cent of water. Through prolonged exposure to the leaching action of meteoric waters, like deposits in the West India Islands have lost all their ammonia salts and other soluble constituents and become converted into insoluble phosphates, or leached guanos like those of the Navassa Islands. (Specimen No. 73243, U.S.N.M., to be noted later; and also specimens from the Grand Con- netables, French Guiana, Nos. 73069 to 73075, U.S.N.M., and Redonda Nos. 53147 to 53152, U.S.N.M.) Origin and occurrence. The origin of the various forms of phos- THE NONMETALLIC MINERALS. 359 phatic deposits has been a subiect of much speculation. Their occur- rence under diverse conditions renders it certain that not all can be traced to a common source, but are the result of different agencies acting under the same or different conditions. By many, all forms are regarded as being phosphatic materials from animal life, and owing their present diversity of form to the varying conditions to which they were at the time of formation or have since been subjected. This, however, as long since pointed out, is an uncalled-for hypothesis, since phosphatic matter must have existed prior to the introduction of animal life, and there is no reason to suppose it may not, under proper conditions, have been brought into combination as phosphate of lime without the intervention of life in any of its forms. The almost universal presence of apatite in small and widely disseminated forms in eruptive rocks of all kinds and all ages would seem to declare its independence of animal origin as completely as the pyroxenic, feld- spathic, or quartzose constituents with which it is there associated. The occurrence of certain of the Canadian apatites as noted later, in veins and pockets, sometimes with a banded or concretionary structure and blending gradually into the country rock, is regarded by some as strongly suggestive of an origin by deposition from solu- tion, that is, by a process of segregation of phosphates from the surrounding rock contemporaneously with their metamorphism and crystallization. Dr. Ells, of the Canadian survey, would regard those occurring in close juxtaposition with eruptive pyroxenites as due to combination of the phosphoric acid brought up in vapors along the line of contact with the calcareous materials in the already softened gneisses. This explanation as well as others will perhaps be better understood in the part of this work relating to localities. On the other hand, the pres- ence of apatite in crystalline form associated with beds of iron ore, as in northern New York, has been regarded by Prof. W. P. Blake and others as indicative of an organic and sedimentary origin for both minerals. The Norwegian apatite from its association with an erup- tive rock (gabbro) has been regarded as itself of eruptive origin. The phosphorites, like the apatites, occur in commercial quantities mainly among the older rocks, and in pockets and veins so situated as to lead to the conclusion that they are secondary products derived by a process of segregation from the inclosing material. Davies regards the Bordeaux phosphorites occurring in the Jurassic limestones of southern France as the result of phosphatic matter deposited on the rocky floor of an Eocene ocean, from water largely impregnated with it. Others have considered them as geyserine ejections, or due to infil- tration of water charged with phosphatic matter derived from the bones in the overlying clays. Stanier, on the other hand, regards the phos- phorites of Portugal as due to segregation of phosphatic matter from the surrounding granite, the solvent being meteoric waters. These 360 EEPOKT OF NATIONAL MUSEUM, 1899. deposits are regarded as superficial and limited to those portions of the I'ock affected by surface waters. The origin of the amorphous, nodular, and massive rock phosphates can, as a rule, be traced more directly to organic agencies. All things considered, it seems most probable that the phosphatic matter itself was contained in the numerous animal remains, which, in the shape of phosphatic limestones, marls, and guanos, have accumulated under favorable conditions to form deposits of very considerable thickness. Throughout these beds the phosphatic matter would, in most cases, be disseminated in amounts too sparing to be of economic value, but it has since their deposition been concentrated by a leaching out by per- colating waters of the more soluble carbonate of lime. Thus H. Losne, in writing of the nodular phosphates occurring in pockety masses in clay near Doullens (France), argues that the nodules as well as the clay itself are due to the decalcification of preexisting chalk by percolating meteoric waters. In this connection it is instructive to note that phosphatic nodules, in size rarely exceeding 4 to 6 cm. , were dredged up during the Chal- lenge?' expedition from depths of from 98 to 1,900 fathoms on the Agualhas Banks, south of the Cape of Good Hope. These are rounded and very irregular capricious forms, sometimes angular and have exteriorly a glazed appearance, due to a thin coating of oxides of iron and manganese. The nodules yield from 19.96 to 23.54 per cent P 2 O 5 . In those from deep water there are found an abundance of calcareous organic remains, especially of rhizopods. The phosphate penetrates the shell in every part, and replaces the original carbonate of lime. The nodules are most abundant apparently where there are great and rapid changes of temperature due to alternating warm and cold oceanic currents, as off the Cape of Good Hope and eastern coast of North America. Under such conditions, together with perhaps altered degrees of salinity, marine organisms would be killed in great num- bers, and by the accumulation of their remains would, it is believed, furnish the necessary phosphatic matter for these nodules. It seems probable that the Cretaceous and Tertiary deposits in various parts of the world may have formed under similar conditions. Hughes has described * phosphatic coralline limestones on the islands of Barbuda and Aruba (West Indies), as having undoubtedly originated through a replacement of their original carbonic by phosphoric acid, the latter acid being derived from the overlying guano. The phos- phatic guano has, however, now completely disappeared through the leaching and erosive action of water, leaving the coral rock itself con- taining 70 to 80 per cent phosphate of lime. Hayes 2 regards the Tennessee black phosphates (Specimens Nos. Quarterly Journal of the Geological Society of London, XLI, 1885, p. 80. "Sixteenth Annual Report of the TJ. S. Geological Survey, 1894-95, Pt. 4, p. (520; Seventeenth Annual Report U. S. Geological Survey, 1895-96, Pt. 2, p. 22. THE NONMETALLIC MINEEALS. 361 62574 and 62781, U.S.N.M.) as due to the slow accumulation on sea bot- toms of phosphatic organisms (Lingulse), from which the carbonate of lime was gradually removed by the leaching action of carbonated waters, leaving the less soluble phosphate behind. The white bedded phosphates of Perry County (Specimen No. 52060, U.S.N.M.), in the same State, are regarded as a product of secondary replacement that is, as due to phos- phate of lime in solution, replacing the carbonate of lime of preexisting limestones, as in the case noted above. The source of the phosphoric acid, whether from the overlying Carboniferous limestones or from the older Devonian and Silurian rocks, is not, however, in this case apparent. Teall has shown 1 that some phosphatic rocks from Clipperton Atoll, in the northern Pacific, are trachytes in which phosphoric acid has replaced the original silica. The replacement he regards as having been effected through the agency of alkaline (principally ammonium) phosphate which has leached down from overlying guano. A micro- scopic examination of the rock in thin sections showed that the replacing process began with the interstitial matter, then extended to the feldspar microlites, and lastly the porphyritic sanidin crystals. The gradual change in the relative proportion of silica and phosphoric acid, as shown lay analyses of more or less altered samples, is shown below, No. I being that of the unaltered rock and II and III of the altered forms: Constituents. I. II. II SiO 2 54.0 43.7 ? P 2 O 5 8.4 17.0 38 Loss on ignition 3.8 12.3 23 J From a comparison of these rocks with those of Redonda, in the Spanish West Indies, it is concluded that the latter phosphates have likewise resulted from a similar replacement in andesitic rocks. (Speci- mens Nos. 53148 to 53152, U.S.N.M.) In this connection reference is made to the work of M. A. Gautier, 2 in which he describes the formation of aluminous phosphates in caves through the action of the ammonium phosphate arising from decomposing organic matter on the clay of the floor of caverns. (See under Occurrences.) The guanos, as noted elsewhere, owe their origin mainly to the accu- mulations of sea-fowl excretions. Such deposits when unleashed, are relatively poor in phosphatic matter and rich in salts of ammonia. Where, however, subjected to the leaching action of rains the more soluble constituents are carried away, leaving the less soluble phos- phates, together with impurities, in the shape of alumina, silica, and iron oxides to form the so-called leached guanos of the West India Islands. As stated in the descriptions of localities, guano deposits are 1 Quarterly Journal of the Geological Society of London, LIV, 1898, p. 230. 2 Formation des Phosphates Naturels d' Alumina et de Fer, Comptes Rendus de 1' Academic des Sciences, Paris, CXVI, 1893, p. 1491. 362 REPORT OF NATIONAL MUSEUM, 1899. not infrequently of a thickness such as to cause their origin as above stated to seem well-nigh incredible were there not sufficient data acquired within historic times to demonstrate its accuracy beyond dis- pute. Thus it is said 1 that in the year 1840 a vessel loaded with guano on the island of Ichabo, on the east coast of Africa. During the excavations which were necessary the crew exhumed the dead body of a Portuguese sailor, who, according to the headboard on which his name and date of burial had been carved with a knife, had been interred fifty-two years previously. The top of this headboard projected 2 feet above the original surface, but had been covered by exactly 7 feet of subsequent deposit of guano. That is to say, the deposition was going on at the rate of a little over an inch and a half yearly. LOCALITIES OF PHOSPHATES. Canada. According to Dr. Ells, of the Canadian Survey, 2 the dis- covery of apatite in the Laurentian rocks of eastern Canada was first made in the vicinitv of the Lievre by Lieutenant Ingall in 1829, though it was not until early in 1860 that actual mining was begun. The mineral occurs in the form of well-defined crystals in a matrix of coarsely crystalline calcite (Specimen No. 67942, U.S.N.M.) and in vein-like and pockety granular masses along the line of contact between eruptive pyroxenites and Laurentian gneisses. The first form is the predominant one for Ontario only, the second for Quebec. From a series of openings made at the North Star Mine, in the region north of Ottawa, it appears that the massive coarsely crystalline gran- ular apatite follows a somewhat regular course in the pyroxenite near the gneiss, but occurs principally in a series of large bunches or chim- neys connected with each other by smaller strings or leaders. Some- times these pockety bunches of ore are of irregular shape and yield hundreds of tons, but present none of the characteristics of veins, either in the presence of hanging or foot walls, while many of the masses of apatite appear to be completely isolated in the mass of pyroxene, though possibly there may have been a connection through small fissures with other deposits. The lack of any connection between these massive apatites and the regularly stratified gneiss is evident, and their occurrence in the pyroxene is further evidence in support of the view th&t these workable deposits are not of organic origin, but confined entirely to igneous rocks. In certain cases where a supposed true-vein structure has been found, such structure can be explained by noticing that the deposits of phosphates occur, for the most part at least, near the line of contact between the pyroxene and the gneiss. By far the greater part of the Canadian apatite thus far mined has been from the Ottawa district of Quebec, where it is mined or quar- ried mainly from open cuts and shafts. The principal fields lie in 1 R. Ridgway, Science, XXI, 1893, p. 360. 2 The Canadian Mining and Mechanical Review, March, 1893. THE NONMETALLIC MINERALS. 363 Ottawa County, Province of Quebec (Specimen No. 62157, U.S.N.M.) and Leeds, Lanark (Specimens Nos. 62136, 62137, U.S.N.M.), Fron- tenac (Specimen No. 62148, U.S.N.M.), Addington, and Renfrew (Specimen No. 62130, U.S.N.M.) counties, Province of Ontario. The first consists of a belt running from near the Ottawa River on the south for over 60 miles in a northerly direction through Buckingham, Portland, Templeton, Wakefield, Denholm, Bowman, Hincks, and other townships to the northward have an average width of 15 to 25 miles. The second belt runs from about 15 miles north of the St. Lawrence River in a northerly direction to the Ottawa River, a distance of about 100 miles, and varies from 50 to 75 miles in breadth. Davies gives the following table as showing the average composition of the Canadian phosphates: Constituents. I. II. III. IV. V. VI. Moisture, water of combination, and loss on ignition . 0.62 33.51 0.10 41 54 0.11 37 68 1.09 30 84 0.89 32 53 1.83 31 87 Lime 46 14 54 74 51 04 42 72 44 26 43 62 7 83 3 03 6 88 13 3? \>> 15 9 28 Insoluble siliceous matter 11.90 0.59 4.29 12.03 10.17 13.50 Equal to tribasic phosphate of lime 100.00 73.15 100.00 90.68 100.00 82.25 100.00 67.32 100.00 71.01 100.10 69.35 Norway. The principal apatite fields lie along the coast in the southern portion of the peninsula between Langesund and Arendal. The material occurs in crystals and crystalline granular aggregates of a white, yellow, greenish, or red color in veins and pockets embedded in the mass of an eruptive gabbro, near the line of contact of the gabbro and adjacent rocks, in the country rock itself in the immediate vicinity of the gabbro, and in coarse pegmatitic veins which are cut by the gabbro. The largest veins are in the mass of the gabbro itself or near the line of contact. Where the apatite occurs in the gabbro the latter is as a rule altered into a hornblende scapolite rock. The principal associated minerals are quartz, mica, tourmaline, scapolite, feldspars, rutile, and magnetic and titanic iron and sulphides of iron and copper. The country rock is gneiss, schist, and granite. The mineral belongs to the variety called fluor apatite, as shown by the following analysis from Dr. Penrose's Bulletin: Apatite from Arendal. Phosphoric acid (P 2 5 ) () 42.229 Fluorine ( 2 ) 3. 415 Chlorine ( 3 ) 0. 512 Lime (CaO) 49. 96 Calcium . . .3. 884 100. 000 1 Equal 92.189 per cent tribasic phosphate. 'Equal 7.01 per cent fluoride of calcium. 'E.qual 0.801 per cent chloride of calcium. 364 REPORT OF NATIONAL MUSEUM, 1899. The Norway apatites have been mined according to Penrose since 1854, the earliest workings being at Kragero. According to Davies, however, the discovery of deposits that could be profitably worked dates only from 1871. The distribution of the material is very uncer- tain and irregular, and the value of the deposits can not be foretold with any great approximation to accuracy. Specimen No. 65122, U.S.N.M., is characteristic. The large specimen on floor of hall, weighing nearly 2 tons, shows well the massive character of the material. A second locality of phosphates but recently described, and which seems to occur under somewhat similar conditions, exists in the Gelli- vara Mountains, in Norrland. Nodular phosphatic deposits are described by Penrose 1 as being found at intervals all along the Atlantic coast of the United States, from North Carolina down to the southern extremity of Florida. The North Carolina deposits occur principally in the counties of Sampson, Duplin, Pender, Onslow, Columbus, and New Hanover, all in the southeastern part of the State. The deposits are of two kinds, (1) a nodular form overlying the Eocene marls and consisting of phosphate nodules, sharks' teeth (Specimen No. 73643, U.S.N.M.), and bones as embedded in a sandy or marly matrix, and (2) as a conglomerate of phosphate pebbles, sharks' teeth, bones, and quartz pebbles, all well rounded and cemented together along with grains of green sand in a calcareous matrix. (Specimen No. 44244, U.S.N.M.) The beds of the first variety usually overlie strata of shell marl, though this is sometimes replaced by a pale green indurated sand. The two following sections will serve to illustrate their mode of occurrence: SAMPSON COUNTY. DUPLIN COUNTY. (1) Soil, sand or clay, 5 to 10 feet. (1) Sandy soil, 1 to 10 feet. (2) Shell marl, 5 to 10 feet. (2) Nodule bed, 1 to 2 feet. (3) Bed with phosphate nodules, 1 to 3 (3) Shell marl. feet. (4) Sea green, sandy marl, 2 to 4 feet. (5) Ferruginous hardpan, 6 to 12 inches. (6) Interstratified lignites and sands as in (4). The nodules as described are of a lead gray color, varying in size from that of a man's fist to masses weighing several hundred pounds. In texture they vary from close compact and homogeneous masses to coarse-grained and highly siliceous rocks distinguished by considerable quantities of sand and quartz pebbles sometimes the size of a chestnut. Occasionally the nodules, which as a rule are of an oval flattened form, 1 Bulletin 46 of the U. S. Geological Survey, 1888. THE NONMETALLIC MINEBALS. 365 contain Tertiary shells (Specimens Nos. 44244 and 34318, U.S.N.M.). The second or conglomerate variety occurs mainly in New Hanover and Fender counties, the beds in some instances being 6 feet in thick- ness, though usually much less. The following section, taken like those above from Dr. Penrose's Bulletin, shows their position and asso- ciation as displayed at Castle Hayne, New Hanover County. (1) AVhite sand, to 3 feet. (2) Brown and red ferruginous sandy clay, or clayey sand, 1 to 3 feet. (3) Green clay, 6 to 12 inches. (4) Dark brown indurated peat, 3 to 12 inches. (5) White calcareous marl, to 2 feet. (6) White shell rock, to 14 inches. (7) Phosphatic conglomerate, 1 to 3 feet. (8) Gray marl containing smaller nodules than the overlying beds, 2 to 4 feet. (9) Light-colored, calcareous marl, containing nodules which are smaller than those in the overlying beds, which grow fewer and smaller at a depth. Many shells. The phosphatic nodules in this conglomerate are kidney and egg shaped as sometimes make up as much as three-fourths the contents of abed; usually, however, the proportion is smaller, and .sometimes there are none at all. The mass as a whole does not contain more than 10 to 20 per cent phosphate of lime, but it is said to have been successfully used as a fertilizer. The individual may be richer in phosphatic mat- ter on the outer surface than toward the center. Aside from the phosphatic layer as described above, phosphatic nodules are found in large quantities in the beds of rivers of these districts, where they have accumulated through the washing action of flowing water, the finer sand clay and gravel having been carried away. Such phosphates naturally do not differ materially from those on land except that they are darker in color and sometimes more siliceous. The deposits of South Carolina, though of low grade compared with some others, are now more generally used than any other known phos- phate. The output of the mines, which is yearly increasing, is shipped to the North, South, and East by sea and to the West by rail. This popularity is due not only to the cheapness of the phosphate ($5 to $6 a ton in 1886), but to the many good qualities of the low-grade acid phosphate made from it. The fact that the nodule bed extends, at an accessible depth, over many miles of country, the easy approach for large vessels up to the very mines, the abundance of water, fuel, and labor, and a climate that permits mining operations to be carried on throughout the whole year, all combine to make the South Carolina phosphates the cheapest and consequently the most productive source of supply of this material. Specimens Nos. 34317 and 34318, 34321 to 34324, and 34326 to 34328, U.S.N.M. are characteristic. Phosphates in the form of nodules and phosphatic marls and green- sands occur in Alabama in both the Tertiary and Cretaceous forma- tions. Their geographical distribution is therefore limited to areas 366 REPORT OF NATIONAL MUSEUM, 1899. south of the outcrops of the lowest Cretaceous beds which stretch in a curve from the northwest corner of the State across near Fayette, Courthouse, Tuscaloosa, Centerville, and Wetumpka, to Columbus, Georgia. As all the Cretaceous and Tertiary beds have a dip toward the Gulf of from 25 to 40 feet to the mile, the phosphate-bearing strata appear at the surface only in a comparatively narrow belt along the line above indicated and are to be found only at gradually increasing depths below at points to the southward. Phosphatic nodules and marls of the Tertiary occur in four different horizons: The Black Bluffs and Nantehala groups of the Lignitic; in the white limestone, and in eastern Alabama, at Ozark, in strata of the Claiborne group. Selected nodules run as high as 27 per cent of phosphoric acid, and marls as high as 6.7 per cent. The Tertiary is not, however, regarded by Professor Smith as a promising source of commercial phosphates in the State. In the Cretaceous the phosphates occur in the transition beds both above and below the so-called Rotten Limestone existing as nodules, shell casts, phosphatic limestones, marls, and greensands. The nodules have essentially the characteristics of those of South Carolina. The principal phosphate region of Florida, as known to-day, com- prises an area extending from west of the Apalachicola River eastward and southward to nearly 50 miles south of Caloosahatchee River, as shown on the accompanying map. 1 According to Mr. Eldridge, the deposits comprise four distinct and widely different classes of commer- cial phosphates, each having a peculiar genesis, a peculiar form of deposit, and chemical and physical properties such as readily distin- guish it from any of the others. According to their predominant characteristics or modes of occur- rence, these classes have come to be known as hard-rock phosphates, soft phosphate, land pebble or matrix rock, and river pebble. With the exception of the soft phosphates, they underlie distinct regions, each class being separate or but slightly commingling with one another. The type of the hard-rock phosphate, as described by Mr. Eldridge, is a hard, massive, close-textured homogeneous, light-gray rock, showing large and small irregular cavities, which are usually lined with second- ary mammillary incrustations of phosphate of lime (Specimens Nos. 66737, 66741, U.S.N.M.), the general appearance being that of the calcareous deposits of the preglacial hot springs of the Yellowstone National Park. There are numerous variations in color and physical characteristics from this type, but which can best be comprehended by a study of the collection. This type carries some 36.65 per cent phosphoric anhydride (P 2 O 5 ). The deposits of the hard-rock phosphate lie in Eocene and Miocene strata, occurring in the first named as a bowlder 1 Preliminary sketch of Phosphates of Florida, by George H. Eldridge. Report of U. S. National Museum, 1899 Me PLATE 20. MAP SHOWING PHOSPHATE REGIONS OF FLORIDA After George H. EMridge. THE NONMETALLIC MINEBALS. 367 deposit in a soft matrix of phosphatic sands, clays, and other material, resulting from the disintegration of the hard rock and constituting the soft phosphates. The deposits underlie sands of from 10 to 20 feet in thickness, and have been penetrated to a depth of 60 feet. The phos- phate deposit proper is white, the bowlders of rounded and irregular outline, varying in diameter from 2 or 3 inches to 10 feet. None of the hard-rock deposits of the Eocene originated in the positions they now occupy. The Miocene hard-rock phosphates, on the other hand, lie in regular bedded deposits in situ, as well as in bowlders. The beds lie horizontal but a few feet below the surface, being covered only by superficial sand. The beds as a rule are but from 4 feet to 5 feet thick. The name soft rock, or soft phosphate, as above indicated, is given to the softer material associated with the hard rock, which in part results from the disintegration of the last named. It is also applied somewhat loosely to any variety not distinctly hard. It therefore varies greatly in color, chemical and physical characteristics, and rarely carries more than 20 to 25 per cent of P 2 O 5 (Specimens Nos. 67304, 67319, 67293, 67296, 67297, U.S.N.M.). The name land-pebble phosphate includes pebble from deposits con- sisting of either earthy material carrying fossil remains, grains of quartz, and pisolitic grains of lime phosphate, or else of a material resembling in texture and other characteristics the hard-rock phos- phate. The individual pebbles vary in size up to that of the English walnut, are normally white, but when subjected to percolating water become dark gray or nearly black. The exteriors are quite smooth and glossy, colors and textures uniform, and average some 30 to 35 per cent P 2 O 5 (Specimen No. 61070, U.S.N.M.). The river-pebble varieties differ from the last mainly in mode of occurrence, being found, as the name would indicate, in the beds of streams, where presumably they have accumulated through the wash- ing away of finer and lighter materials. The}' are most abundant in the Peace, Caloosahatcb.ee, Alafia, and other rivers entering the Gulf south of Tampa and Hillsborough bays, though the Withlacoochee, Aucilla, and rivers of the western part of the State, carry also a mix- ture of pebbles, hard-rock fragments, and bones derived from the vari- ous strata through which they have cut their channels. The pebbles of the Western rivers show a very uniform composition, and range from 25 to 30 per cent phosphoric anhydride (P 2 O 5 ), or about 65 per cent of phosphate of lime, the impurities being mainly siliceous matter, car- bonate of lime, alumina, and iron oxides (Specimens Nos. 67299, 67298, 67355, U.S.N.M.). Phosphates the mineralogical nature of which does not seem to be as yet accurately made out occur in the Devonian Shales of Middle Tennessee. They are thus described by Professor Safford: 1 Engineering and Mining Journal, LVII, April 21, 1894, p. 366. 368 REPOKT OF NATIONAL MUSEUM, 1899. There are two distinct beds of the phosphates, one above a stratum known as the black shale; the other below the shale. The one ubovo is u bed or layer of concretionary masses, balls, and kidney and knee-shaped forms from the size of walnuts to that of a man's head (Specimen No. 52059 from Hickman County). These are sometimes loosely disposed in a greenish or bluish shale, and sometimes tightly packed together like so many cannon balls in a layer 8 or 10 inches thick. Ordinarily the layer has less thickness, often, in fact, being represented by only a few scattered concretions. But, thick or thin, it may be said to be universally present, its kidneys serving to indicate the place of the black shale and the underlying bed when these are concealed by debris or soil. The other phosphate, that underlying the shale, and the more impor- tant of the two, is, in its best presentations, a well-defined, continuous stratum of dark-bluish or bluish-black rarely grayish rock, with fine or coarse grain. Its regularly stratified character and its dark color make it look like a bed of stone coal. The geographical distribution and general geology of these phos- phates has been worked out in considerable detail by C. W. Hayes, to whose papers reference has been already made (p. 360). Accord- ing to this authority the phosphates occur in four distinct varieties: (1) Black nodular phosphate; (2) black bedded phosphate; (3) white breccia phosphate, and (4) white bedded phosphate. The first two of these are of Devonian age, the third is a secondary and comparatively recent deposit, while the fourth, perhaps also of secondary origin, is interbedded with rocks of Carboniferous age. The black nodular variety contains from 60 to 70 per cent of phosphate of lime, and is found in commercial quantities only in the region of the black bedded phosphate in western middle Tennessee. The black bedded variety, which is the only one that has thus far proved of commercial importance, is confined, so far as at present known, "to an oval area southwest of Nashville, having Centerville about in its center." It also covers portions of Hickman, Williamson, Maury, Lewis, Wayne, Perry, and Decatur counties. Sections showing the relation of the phosphates to the adjacent for- mations are given in Dr. Hayes's paper. The beds vary in thickness from a fraction of 1 to 8 or 10 feet, the average run of the rock being about 50 per cent phosphate of lime. The white bedded and white breccia phosphates are limited to small areas in Perry County. Their contents of phosphoric acid (P 2 O 5 ) is low, varying from 14 to 15 per cent, and as yet their value for other than local purposes is to be deter- mined. (See especially Specimens Nos. 52058, 52060, 52061, U. S. N. M. ) England. Deposits of phosphates sufficiently concentrated for com- mercial purposes lie near the upper limit of Cambro- Silurian strata in North Wales. According to Davies, the phosphatic material occurs THE NONMETALLIC MINERALS. 369 in the form of nodular concretions of a size varying from that of an egg to a cocoanut, closely packed together and cemented by a black slaty matrix. The concretions have often a black highly polished appearance, due to the presence of graphite, but owing to the pres- ence of oxidizing pyrite they sometimes become rusty brown. The concretions carry from 60 to 69 per cent of phosphate of lime; the matrix is also phosphatic. The phosphate beds are highly tilted and are overlaid by gray shales with fossilized echinoderms and underlaid by dark crystalline limestone, which also contains from 15 to 20 per cent of phosphatic material. Davies regards the deposit as an old sea bottom "on which the phosphatic matter of Cretaceous and Molluscan life was precipitated and stored during a long period, while certain marine plants may also have contributed their share of phosphatic matter. He thinks it also as possible that, as in the Lauren tian deposits, the water of the sea may have contained phosphatic matter in solution, to be deposited independently of organic agencies. These phosphated beds are mined at Berwin, where an average pro- duction over a space of 360 fathoms was 2 tons 10 hundredweight of phosphate per fathom, of an average strength of 46 per cent. The nodules average from 45 to 55 per cent of phosphate of lime. Amorphous nodular phosphates also occur in both the Upper and Lower Greensands of the Cretaceous and in Tertiary deposits. Those of the upper beds have been mined in Cambridgeshire and Bedford- shire. The phosphatic material occurs in the form of shell casts, fossils, and nodules, of a black or dark-brown color, of varying hard- ness, embedded in a sand consisting of siliceous and calcareous matter as well as phosphatic and glauconitic grains. The average composi- tion shows from 40 to 50 per cent of phosphate of lime. The thick- ness of the nodule-bearing bed is rarely over a foot. The nodules of the Lower Greensands differ from those of the Upper in many details, the more important being their lower percentages of phosphate of lime (from 40 to 50 per cent). They occur in a bed of siliceous sand which itself is not phosphatic. The Tertiary phosphates reach their best development in the county of Suffolk, where they are found at the base of the Coralline and Red Crog groups and immediately over- lying the London clays. The beds consist of a "mass of phosphatic nodules and shell casts, siliceous pebbles, teeth of cretaceans and sharks, and many mammal bones, besides occasional fragments of Lower Greensand chert, granite, and chalk flints." The nodules vary in both quality and quantity. They are at times of a compact and brittle nature, while at others they are tough and siliceous. They average about 53 per cent phosphate of lime and 13 per cent phosphate of iron. France. Phosphates of the nodular type occur in beds of Cretaceous age in the provinces of Ardennes and Meuse, and to a less extent in others in northern France; in the department of Cote d'Or, and along NAT MUS 99 24 370 REPORT OF NATIONAL MUSEUM, 1899. the Rhone at Bellegarde, Seyssel, and Grenoble. As in England, the phosphatic nodules of the northern area, such as are of commercial importance, occur in both the Upper and Lower Greensands. They resemble in a general way the English phosphates, but are described as soft and porous and easily disintegrating when exposed to the air. Those of the Upper Greensand average some 55 per cent of phosphate of lime. More recently deposits have been described by M. J. Gosselet, 1 near Fresnoy-le-Grand, in the north of France. The phosphatic material occurs in a zone of gray chalk some 6 feet in thickness (1 to 2 meters), and is in the form of concretionary nodules forming a sort of con- glomerate in the lower part of the bed. A portion of the chalk is also phosphatic. Phosphatic material (of the type of phosphorites) is found in fissures and pockets in the upper portion of limestones of Middle Jurassic (Oxfordian) age, in the departments of Tarn-et- Garonne, Aveyron, and Zoti, France. The deposits are of two kinds. The first occurring in irregular cavities or pockets never over a few yards long, and the second in the form of elongated leads with the sides nearly vertical. These are generally shallow, and thin out very rapidly at a short distance below the surface. The nodules or concretions are of a white or gray color, waxy luster, and opal-like appearance, and occur in the form of tubercular or kidney shaped masses embedded in ferruginous clay in the clefts of the lime- stone, or in geodic, fibrous, and radiating forms. The material of this region is known commercially as Bordeaux phosphate, being shipped mainly from Bordeaux. They average from 70 to 75 per cent phosphate of lime, the impurities being mainly iron oxides and siliceous matter. Gautier 2 describes deposits of phosphates estimated to the amount of 120,000 to 300,000 tons on the floors of the Grotte de Minerve, near the village of Minerve on the northeast flank of the Pyrenees, in Aude, France. The cave proper is in nummulitic limestone of Eocene age, the floors being formed by Devonian rocks. The filling material con- sists of cave earth and bone breccia below which are the aggregates of concretionary phosphorites and other phosphatic compounds of lime and alumina, the more interesting being Brushite^ a hydrous tribasic calcium phosphate hitherto known only as a secondary incrustation on guano from the West India islands, and Minervite, a new species hav- ing the formula A1 2 O 3 . P 2 O 5 , 7H a O, a hydrous aluminum phosphate, existing in the form of a white plastic clay-like mass filling a vein from a few inches to 2 or more feet in thickness. Germany. According to Da vies, the principal phosphate regions of 1 Annales de la Societe Geologique du Nord., XXI, 1893, p. 149. 2 Annals des Mines, V, 1894, p. 5. THE NONMETALLIC MINERALS. 37 1 North Germany occupy an irregular area bounded on the northeast by the town of Weilburg, on the northwest by the Westerwald, on the east by the Taunus Mountains, and on the south by the town of Dietz. The phosphorite occurs in the form of irregular nodular masses of all sizes up to masses of several tons weight, embedded in clay which rests upon Devonian limestone and is overlaid by another stratum of clay. The phosphate-bearing clay varies in thickness from 6 inches to 10 feet. With the phosphate nodules are not infrequently associated deposits of manganese and hematite. Davies regards the deposits as of early Tertiary age. The color of the freshly mined material varies from pale buff to dark brown, varying in specific gravity from 1.9 to 2.8, the quality deteriorating with the increase in gravity. Selected samples of the staple nodules yielded as high as 92 per cent phosphate of lime; but the average is much lower, being but about 50 to 60 per cent phosphate of lime. (Specimens Nos. 66827, 66828, U.S.N.M., from Gleisenberg and Heckholzhausen.) Belgium. Nodular phosphates belonging to the Upper Cretaceous formations occur in the province of Hainaut, where they form the basis of an extensive industry. The nodules, which are generally of a brown color and vary in size from the fraction of 1 to 4 or 5 inches in diameter, lie in a coarse-grained, friable rock called the brown or gray chalk, which itself immediately underlies what is known as the Ciple} 7 conglomerate. The phosphate-bearing bed is sometimes nearly 100 feet in thickness, but is richest in the upper 10 feet, where it is estimated the phosphatic pebbles constitute some 75 per cent of its bulk. Below this the bed grows gradually poorer, passing by grada- tions into the white chalk below. The overlying conglomerate also carries phosphate nodules, which carry from 25 to 50 per cent phosphate of lime. Owing to the hard- ness of the inclosing rock they are less mined than those in the beds beneath. The mining of phosphates is carried on extensively near the town of Mons, on the lands of the communes of Cuesmes, Ciply, Mes- vin, Nouvelles, Spiennes, St. Symphorien, and Hyon. The annual output has gradually increased from between 3,000 and 4,000 tons in 1887 to 85,000 tons in 1894. Other phosphatic deposits are described 1 as occurring in the provinces of Antwerp arid Liege. Spain. Important deposits of phosphorites occur between Logrosan and Caceres, in Estremadura Province. The deposits are in the form of pockets and veins in slates and schists supposed to be of Silurian age; at times a vein is found at the line of contact between the slate and granite. The veins vary in thickness from 1 to several feet, the largest being some 20 feet and extending for over 2 miles. This is by 1 Annales de la SociSte G/,! IKJ coals are so called from the facility with which they may be made to yield coke. Such give a yellow flame in burning and make a hot fire. (Specimens Nos. 55490, U.S.N.M., Connellsville, Pennsylvania, and 59260, U.S.N.M., from New River, West Virginia.) Other varie- ties of apparently the same composition and general physical proper- ties can not for some unexplained reason be made to yield coke, and are known as noncoking coals. (Specimens Nos. 59428, U.S.N.M., from Vigo County, Indiana, and 59208, U.S.N.M. (splint coal), from Fayette County, West Virginia.) Cannel coal has a very compact structure, breaks with a conchoidal fracture, has a dull luster, ignites easily, and burns with a yellow flame. It does not coke. Its chief characteristic is the large amount of volatile matter given off when heated, whereby it is rendered of particular value for making gas. (Specimens Nos. 56280, 56284, and 58496, U.S.N.M., are characteristic.) Before the discovery of petroleum it was used for the distillation of oils. Below is given the composition of a (1) coking coal from the Geology of England and Wales, p. 278. Report of U. S. National Museum, 1899 Mar PLATE 24. c ." THE NONMETALLIC MINERALS. 427 Connellsville Basin of Pennsylvania, and (II) a cannel coal froiu Ka- nawha Countj 7 , West Virginia. 1 Constituents. I. II. Water 1 105 Volatile matter 29.885 58.00 Fixed carbon Ash 57.754 9 895 23.50 18 50 Sulphur 1.339 Total 99.978 100.00 ANTHRACITE COAL. This is a deep black, lustrous, hard and brittle variety, and represents the most highly metamorphosed variety of the coal series. Traces of organic nature are almost entirely lacking in the matter of the anthracite itself, though impressions of ferns, lyco- pods, sigillaria and other coal-forming plants are frequently associated with the beds in such a manner as to leave little doubt as to their origin. Anthracite is ignited with difficulty and burns with little flame, but makes a hot fire. Below is given the average composition of a coal from the Kohinoor Colliery, Shenandoah, Pennsylvania. 8 Water 3. 163 Volatile matter 3. 717 Fixed carbon 81. 143 Sulphur 0. 899 Ash . 11.078 100. 00 (Specimens Nos. 59058, 59062, from Pennsylvania, and 30854, from Colorado, are sufficiently characteristic.) Like the other coals, anthra- cite occurs in true beds, but is confined mostly to rocks of the Car- boniferous age. Thin seams of anthracite sometimes occur in Devo- nian and Silurian rocks, but which are too small to be of economic value. Rarely coals of more recent geological horizon have been formed locally, altered into anthracite hy the heat of igneous rocks. Through a still further metamorphism, whereby it loses all its volatile constituents, coal passes over into graphite (Specimens Nos. 17299 and 59099, from near Newport, Rhode Island), and it is possible, though scarcely probable, that all graphite may^ have originated in this way. The principal anthracite coal regions of the United States are in eastern Pennsylvania. From here westward throughout the interior States to the front range of the Rocky Mountains the coals are all soft, bituminous coals. Those of the Rocky Mountain region proper are largely lignitic, passing into the bituminous varieties. 1 F. P. Dewey, Bulletin 42, United States National Museum, 1891, p. 231. 1 Idem, p. 221. 428 REPORT OF NATIONAL MUSEUM, 1899. BIBLIOGRAPHY. The bibliography of coal, even though limited to the United States, would be enor- mous. In all cases reference should be made to the publications of the various State surveys, where such have existed. The few titles here given are of articles of general interest and, as a rule, not relating to the coals of one particular locality alone. WALTER R. JOHNSON. A Report to the Navy Department of the United States on American Coals Applicable to Steam Navigation and to other purposes. Washington, D. C., 1844, pp. 607. RICHARD COWLING TAYLOR. Statistics of Coal. The Geographical and Geological Distribution of Mineral Combustibles or Fossil Fuel, etc. Philadelphia, 1848, pp. 754. J. LE CONTE. Lectures on Coal. Report of the Smithsonian Institution, 1857, p. 119. T. H. LEAVITT. Peat as a Fuel. Second Edition. Boston, 1866, pp. 168. Facts About Peat as an Article of Fuel. Third Edition. Boston, 1867, pp. 316. E. W. HILGARD. Note on Lignite Beds and their Under Clays. American Journal of Science, VII, 1874, p. 208. LEO LESQUEREUX. On the Formation of Lignite Beds of the Rocky Mountain Region. American Journal of Science, VII, 1874, p. 29. J. S. NEWBERRY. On the Lignites and Plant Beds of Western America. American Journal of Science, VII, 1874, p. 399. JAMES MACFARLANE. Coal Regions of America. New York, 1875. MIALL GREEN, THORPE, RUCKER, and MARSHALL. Coal; Its History and Uses. Edited by Professor Thorpe. London, 1878, pp. 363. RAPHAEL PUMPELLY. Report on the Mining Industries of the United States, with special investigation into the Iron Resources of the Republic and into the Creta- ceous Coals of the Northwest, Tenth Census of the United States, XV, 1880. W. IVISON MACADAM. Analyses of Coals from New Zealand and Labuan. Transactions of the Edinburgh Geological Society, IV, pt. 2, p. 165, session 1881-82. J. S. NEWBERRY. On the Physical Conditions under which Coal was Formed. Science, I, March 2, 1883, p. 89. CHARLES A. ASHBURNER. The Classification and Composition of Pennsylvania Anthra- cites. Transactions of the American Institute of Mining Engineers, XIV, 1885, p. 706. LEO LESQUEREUX. On the Vegetable Origin of Coal. Annual Report of the Geological Survey of Pennsylvania, 1885, p. 95. S. W. JOHNSON. Peat and its Uses as Fertilizer and Fuel. New York, 1886, pp. 168. GRAHAM MACFARLANE. Notes on American Cannel Coal. Transactions of the American Institute of Mining Engineers, XVIII, 1890. p. 436. W. GALLOWAY. The South African Coal Field. Proceedings of the South Wales Institute of Engineers, No. 2, XVII, 1890, p. 67. LEVI W. MEYERS. L'Origine de la Houille. Revue de Quest. Scientifique Brussels, July, 1892, pp. 5-47. WILLIAM H. PAGE. The Carboniferous Age and the Origin of Coal. Engineering and Mining Journal, LVI, 1893, p. 347. Note sur la formation des Terraines Houillers. Bulletin de la Societe" Geologique de France, XXIV, 1896, p. 150. Making Coal of Bog Peat. The Iron Age, LXII, Aug. 18, 1898, p. 3. Report of U. S. National Museum, 1899. Merrill. PLATE 25. m - : ^ ~-~^AJ^ - x ' " * ' '/ il t ./ 2 S => u i THE NONMETALLIC MINERALS. 429 Bituminous . 2. BITUMEN SERIES. Under this head are included a series of hydrocarbon compounds varying in physical properties from solid to gaseous and in color from coal black through brown, greenish, red, and yellow to colorless. Unlike the members of the series already described, they are not the residual products of plant decomposition in situ, but are rather, in part at least, distillation products from deeply buried organic matter of both animal and vegetable origin. The different members of the series differ so widely in their properties and uses that each must be dis- cussed independently. The grouping of the various compounds as given below is open to many objections from a strictly scientific stand- point, but, all things considered, it seems best suited for the present purposes. 1 Tabular classification of hydrocarbons.'* Gaseous Marsh gas (Natural gas). Fluidal Petroleum (Naphtha). .,. fPittasphalt (Maltha). \iscous and sem.sohd Minera] ^ I Asphalt, Elastic (Elaterite. \Wurtzillite. fAlbertite. 8011(1 Grahamite. 'Uintaite. Succinite. Resinous.. Copalite. Torbanite. Ambrite. Cerous ( waxy) . . . f Ozokerite. "IHatchettite. Tabular classification or grouping of natural and artificial bituminous compounds. Mixed with limestone, "asphal- fSeyssel, Val de Travers, Lobsan, Illi- tic limestone." I nois, and other localities. Mixed with silica and sand, "as- f California, Kentucky, Utah, and other phal tic sand." I localities. "Bituminous silica." Mixed with earthy matter, "as- f phaltic earth " ^ Trinidad, Cuba, California, Utah. (.Bituminous schists... f Canada, California, Kentucky, Virginia, \ and other localities. (Thick oils from the distillation of petro- "l leum. "Residuum." Viscous (Gas-tar. IPitch. Solid Refined Trinidad tic of asphaltite. Gritted asphaltic pounds. asphaltic earth. Mas- mastic. Paving com- 'See article What is Bitumen? by S. F. Peckham, Journal of the Franklin Insti- tute, CXL, 1895, pp. 370 to 383. 2 W. P. Blake, Transactions of the American Institute of Mining Engineers, XVIII, 1890, p. 582. 430 REPORT OF NATIONAL MUSEUM, 1899. Important natural bitumens. Table of occurrence of important natural bitumen.* Natural gas Ohio, Pennsylvania, California, etc., in the United States; Russia, France, etc. Natural naphtha Found in petroleum districts (of little value, superseded by artificial naphtha from crude petroleum). Petroleum Pennsylvania, Ohio, Wyoming, California, etc., in United States; Russia, etc. (consult books on petroleum). Maltha California, Wyoming, Alabama, Utah, Colo- rado, Kentucky, New Mexico, Ohio, Texas, Indian Territory, etc.; Russia, France, Germany, etc. North America Utah, California, Texas, etc. Jentral America. . .Cuba, Mexico, etc. South America Trinidad, Venezuela, Peru, Colombia, etc. Europe Caucasia, Syran-on-the Volga, Germany, France, Italy, Austria, etc. Asia Hit on the Euphrates, Asia Minor, Palestine, etc. Africa Oran in Egypt; probably other places. S^orth America Asphaltum . Asphaltum almost pure. West Virginia, Kentucky, Texas, Wyoming, Utah, Colorado, California, Indian Territory, Mon- tana, New Mexico. Central America. . .Mexico, Cuba, etc. South America Trinidad (largest supply, most used), Venezuela, Asphaltic Peru, Colombia, etc. compounds. Europe Germany, Switzerland, France, Italy, Sicily, Russia, Austria, Spain, etc. Asia Asia Minor, Palestine, Bagdad, and probably in China. Africa Egypt, and probably else- where in Africa. Origin. Of the many views, mainly theoretical, that have been put forward to account for the origin of bituminous compounds, but two need be noted in detail here. Interested readers are referred to the bibliography given on page 460, and particularly to the works of Peckham, Orton, and Redwood. Prof. Edward Orton, after an X J. W. Howard, as quoted by S. P. Sadtler, Journal of the Franklin Institute, CXL, 1895, p. 200. THE NONMETALLIC MINERALS. 431 exhaustive consideration of the occurrence of petroleum, natural gas, and asphalt in Kentucky, 1 gives the following precise summary: 1. Petroleum is derived from organic matter. 2. Petroleum of the Pennsylvania type is derived from the organic matter of bitumi- nous shales, and is probably of vegetable origin. 3. Petroleum of the Canadian type is derived from limestones, and is probably of animal origin. 4. Petroleum has been produced at normal rock temperatures (in American fields), and is not a production of destructive distillation of bituminous shales. ' 5. The stock of petroleum in the rocks is already practically complete. Hofer 2 regards petroleum as of animal origin only, and advances the arguments given below in support of his theory: 1. Oil is found in strata containing animal, but little or no plant remains. This is the case in the Carpathians, and in the limestone examined in Canada and the United States by Sterry Hunt. 2. The shales from which oil and paraffin were obtained in the Liassic oil shales of Swabia and of Steirdorf, in Styria, contained animal, but no vegetable remains. Other shales, as, for instance, the copper shales of Mansfield, where the bitumen amounts to 22 per cent, are rich in animal remains and practically free from vegetable remains. 3. Rocks which are rich in vegetable remains are generally not bituminous. 4. Substances resembling petroleum are produced by the decomposition of animal remains. 3 5. Fraas observed exudations of petroleum from a coral reef on the .shores of the Red Sea, where it could be only of animal origin. The relationship which exists between the solid or viscous bitumen and the fluidal petroleum have not in all cases been satisfactorily worked out, though Peckham has shown 4 that in California at least there are almost infinite gradations from one extreme to the other. In Ventura County, for instance, the petroleum is held, primarily, in strata of shale, from which it issues as petroleum or maltha, accordingly as the shales have been brought into contact with the atmosphere, the asphaltum being produced by a still further exposure to the atmosphere after the bitumen has reached the surface. This relationship between the more fluidal and viscous varieties is shown in tig. 18, copied from Professor Peckhaur s paper above referred to, and which represents a section across a portion of Sulphur Mountain between the Hayward Petroleum Company's tunnels in Wheeler Canyon, and the Big Spring Plateau on the Ojai ranch. In this section it will be noted that the mountain is formed of a synclinal fold of shale, the strata dipping 'Report on the Occurrence of Petroleum, etc., in Western Kentucky. Geological Survey of Kentucky, John R. Proctor, director, 1891. 2 As quoted by Redwood, I, p. 238. 3 Dr. Engler, as quoted by Redwood, obtained by distillation of menhaden oil, among other products, a substance remarkably like petroleum, and a lighting oil indistinguishable from commercial kerosene. 4 See Report of the Tenth Census, p. 68. 432 REPORT OF NATIONAL MUSEUM, 1899. THE NONMETALLIC MINEEAL8. 433 inward on both sides and coming to the surface almost vertically on the right, and more nearly horizontally on the left (the south). The tunnels are driven into the nearly vertical face of the mountain and the oil-bearing rock is protected by some 700 or 800 feet of overlying shales. The oil obtained is the lightest thus far found in southern California. On the other hand, the material which exudes on the north side, when the shales are upturned at such an angle as to give free access to atmospheric agencies, is in the form of maltha, or min- eral tar, and so viscous, in December, 1865, that it could be gathered and rolled into balls like dough. The relationship between petroleum and natural gas is scarcely better defined. That the gas can be derived from petroleum is undoubted, and indeed the latter apparently never occurs free from gas. But on the other hand, as Professor Orton states, the gas often originates under many conditions in which petroleum does not occur. The formation of marsh gas from decomposing plant remains on the bottom of stagnant pools, and its presence in coal mines would show with seeming conclusiveness that a part, at least, of the gas is formed quite independently of petroleum. It would seem on the whole most probable that no one theory was universally applicable to all cases. MARSH GAS: NATURAL GAS. This is a colorless and odorless gas arising from the decomposition of organic matter protected from the oxidizing influence of atmospheric air. By itself it burns quietly, with a slightly luminous flame, but when mixed with air it forms a dangerous explosive. It is this gas which forms the dreaded tire damp of the miners. In small quantities this gas may be found and collected, if desired, from the bottom of shallow pools and stagnant bodies of water by merely disturbing the decomposing plant matter at the bottom, when the bubbles of the gas will rise to the top. Under this head may property be considered the so-called natural ".' > ~****$J^-*^ ^ ^f^3 THE NONMETALLIO MINERALS. 469 files and points. Dentists use particularly the "knife blade," a very thin, broad slip of stone, triangular in section, with one short side, the other two forming a thin edge as they come together (Specimens Nos. 38998, 53721, U.S.N.M.). They are used for filing between the teeth. Carvers use wedge-shaped, flat, square, triangular, diamond-shaped, rounded, and bevel-edged files for finishing their work. (Specimen No. 38996, U.S.N.M.). Jewelers, especially manufacturing jewelers and watchmakers, use all these forms of files and also points. These last are sometimes made the size of a lead pencil, having a cone-shaped end, and are about 3 inches long and i inch square, tapering to a point. They are used chiefly in manufacturing watches to enlarge jewel holes (Specimens Nos. 38995, 53726-53727, U.S.N.M.). Wheels of various thicknesses and diameter are also made from Arkansas stone. They are used chiefly by jewelers and dentists, but could be made of service in all workshops where an Arkansas whet- stone is used (Specimens Nos. 38992, 38962, 53710, U.S.N.M.). The difficulty of obtaining pieces of clear stone large enough for wheels several inches in diameter makes the price very high, and the difficulty of cutting out a circular form increases the cost. Wheels are quoted at from $1.10 to $2.20 an inch of diameter. Arkansas stone is used for finishing and polishing metal rolls, jour- nals, cross-head slides, piston rods, crank pins, and all kinds of lathe work. Fragments of the Arkansas stone are saved at the factories, and now and then sent away to be ground for polishing powder. In the manu- facture of this powder millstones are worn out so rapidly that the process is rather expensive, but as waste stone is utilized, the powder can be sold by the barrel at 10 cents a pound. It makes a very fine, pure white powder of sharp grit, suitable for all kinds of polishing work; it is known as "Arkansas powder." A large amount of energy is wasted, however, in the manufacture of this powder, for the silica of the Ouachita stone is in every way identical with that of the Arkan- sas stone, and it would be much more easily reduced to powder than the Arkansas. The so-called Turkish oilstone from Smyrna, in Asia Minor, is both in structure and abrasive qualities quite similar to the Arkansas novac- ulites. (Specimens Nos. 38956, 38967, 38997, U.S.N.M.) It, however, is of a drab color and carries an appreciable amount of free calcium carbonate and other impurities, as shown by the analysis given below, as quoted by Griswold: TURKEY STONE. Silica (SiO 2 ) 72.00 Alumina (A1,O S ) 3. 33 Lime (CaO) 13.33 Carbonic acid (CO,) 10.33 470 REPORT OF NATIONAL MUSEUM, 1899. According to Renard, 1 the celebrated Belgian razor hone quarried at Lierreux, Sart, Salm-Chateau, Bihau, and Recht is a damourite slate containing innumerable garnets, more than 100,000 in a cubic millimeter. Like the Ratisbon hone, this occurs in the form of thin, yellowish bands, some 6 centimeters wide (2f inches) in a blue-gray slate (phyllade). The bands are essentially parallel with one another and with the grain of the slate, into which they at times gradually merge. The chemical composition of a sample from Recht is given as below. The microscopic structure of the stone as described and figured by Renard is essentially the same as that of the Ratisbon stone in the collection of the IT. S. National Museum (see Plate 30, fig. 2), and the stones are practically identical in color and texture as well. Silica (SiO 2 ) 46.5 Titanic oxide (TiO 2 ) 1.17 Alumina (A1 2 O 3 ) 23. 54 Ferric iron (Fe 2 O s ) 1 . 05 Ferrous iron (FeO) 0. 71 Manganese oxide (MnO) 17.54 Magnesia (MgO) 1.13 Lime (CaO) 0. 80 Soda (Na/)) 0.30 Potash (K 2 O) :.... 2.69 Water (H 2 O) 3.28 Carbon dioxide (CO 2 ) 0. 04 Phosphoric acid (P 2 5 ) 0. 16 Sulphur (S) 0. 18 Organic matter 0. 02 Total 99. 11 The cutting property of the stone would appear to be due to the presence of the small garnets above noted. (Specimens Nos. 38938- 38940, U.S.N.M.) The so-called holystone is but a fine, close-grained sandstone of the same nature as that used in grind and whet stones. The greater part of those made in this country are from the Berea sandstone of Ohio, though some are said to be imported from Germany. The stones are used mainly on shipboard, and the trade is small. 2. PUMICE. The material to which the name pumice is commonly given is a form of glassy volcanic rock, which, by the expansion of its included moist- ure while in a molten condition, has become, like a well-raised loaf, filled with air cavities or vesicles. The cutting or abrasive quality 1 Memoires Couronnes et Memo! res des Savants Etrangers de L' Academic Royal des Sciences, etc., Belgique 1878, pp. 1-44. Report of U. S. National Museum, 1899. M PLATE 30. Fig. 2. MlCROSECTIONS SHOWING THE APPEARANCE OF (1) ARKANSAS NOVACULITE AND (2) RATISBON RAZOR HONE. THE DARK BODIES IN (2) ARE GARNETS. THE NONMEf ALLIC MINERALS. 471 of the material is due to the thin partitions of glass composing the walls between these vesicles. Any variety of volcanic rock, flowing out upon the surface of the ground, is likely to assume the vesicular condition known as pumiceous, but only certain acid varieties known as liparites seem to possess just the right degree of viscosity to produce a desirable pumice, and in this rock only in exceptional circumstances. Almost the entire commercial supply of pumice is now brought from the Lipari Islands, a group of volcanoes north of Sicily, in the Mediterra- nean Sea. (Specimen No. 6078T, U.S.N.M.) The material is usually brought over in bulk and sold in small pieces in the drug and paint shops, or ground and bolted to various degrees of fineness and sold like emery and other abrasive materials. (Specimen No. 54155, U.S.N.M.) At times an inferior grade of pumice has been produced from volcanic flows near Lake Merced, in California. In Harlan County, Nebraska, and adjacent portions of Kansas, as well as in many other of the States and Territories farther west, have been found extensive beds of a fine. white powder, which was first shown by the present writer 1 to be pumiceous dust, drifted an unknown distance by wind currents and finally deposited in the still waters of a lake. Through a mistaken notion regarding its origin this material was first described in Nebraska as yeyserite. So far as the writer is aware, these natural pumice powders have thus far been used only locally for polishing purposes and as a cleansing or scouring agent in soap. As the material exists in almost inexhaustible quantities, it would seem that a wider scope of usefulness might yet be discovered. (Specimens Nos. 53074, 00920. 37023, U.S.N.M., from Montana, Washington, and Nebraska.) The analyses given below show (I) the composition of the pumice dust of Harlan, Orleans County, Nebraska, 2 and (II) a pumice from Capo di Costagna, Lipari Islands: Constituents. I. II. Silica 69.12 73.70 12.27 1 17.64 2.31 0.86 0.65 0.24 0.29 6.64 4.73 1.69 4.25 4.05 1.22 Total 100.24 99.42 1 See On Deposits of Volcanic Dust in Southwestern Nebraska (Proceedings U. S. National Museum, VIII, 1885, p. 99), and Notes on the Composition of Certain Plio- cene Sandstones from Montana and Idaho (American Journal of Science, XXXII, 1886, p. 199). 2 Rocks, Rock-weathering, and Soils, p. 350. 472 REPORT OF NATIONAL MUSEUM, 1899. According to Dr. L. Sambon, as quoted by Dr. H. J. Johnston -Lavis: All the best pumice of commerce is obtained from the northeast region of the island of Lipari, extending as far as the summit of Mte. S. Angelo on its northern slope. * * * It is excavated at the Fossa Castagna near M. Pelato, at M. Chirica, and on the shore of the Mosche. I visited a quarry of M. Pelato on the outer southern side. The height was about 1.50 m. and 1 m. large. The entrance was sustained by poles, faggots of brushwood, and stones; at first one descended for 160 steps, then one ascended for about 50 m. where two naked men were digging in the dull light of an oil lamp. In decending I met some young men who were carrying up baskets full of pumice. They wore short coarse linen drawers, and on their naked breast hung the blessed scapulary. On my arrival at the workes they made me sit down on an empty basket while I watched the men dig out the pieces of pumice, often the size of a human head, from the embedding matrix, which is composed of different sized fragments and dust of the same material, pressed together, and forming an incoherent tuff. They told me of their poor wages, and the dangers of their work in consequence of the frequent collapse of the workings, killing men and youths. It was horrible to hear those accounts of misery and misfortune at the bottom of these caves. The low roof and narrow passage from which every moment fragments detached themselves seemed to threaten the collapse of the whole; and it was with great relief that I again reached the daylight. Only a few weeks previously a quarry of M. Pelato had collapsed and buried some workmen, and more than two days work were required to reach them. These unfortunate men, saved by a miracle, returned again to their work, for what else could they have done to obtain bread? Prolonged and curious was at all times the discussion concerning the origin of pumice. It was believed to be amianthus decomposed by fire, by Pott, Bergman and Demeste; calcined lignite or schist, by Vallerio; scorified marl by Sage and granite that had become blown up and fibrous by the effect of fire and water by Dolomieu. The latter asserted having found inclosed in some pieces of pumice frag- ments of granite. He also declares that he had seen masses of granite which took on gradually the fibrous structure and other characters of pumice; so that he concluded that granite or granitoid schist was the primitive material which by the effect of the volcanic fire passed to the state of the piimice. Finally he declares he sent speci- mens to all the most learned geologists of the time. Spallanzani, who visited that same locality and hunted in every part of Campo Bianco in a most diligent manner but without being successful in finding the granite of Dolomieu, says wittily that probably the French geologist had carried them all away. Spallanzani himself, on the contrary, considers that pumice and obsidian are the result of fusion of great masses of intermediate lavas which one encounters on all parts of the mountain. Prof. J. F. Blake recently, probably ignoring the observations of Spallanzani, is sat- isfied in finding in that locality "Mother-pumice" as he has baptized it, from which also is derived the obsidian. But pumice, obsidian and all intermediate rock varie- ties more or less scoriaceous are but different forms of the same eruptive product. The whole history and modifications of pumice have been worked out by Dr. John- ston-Lavis, who has shown that by studying these eruptive products the whole mechanism of volcanic action in general is explained and the sequence of eruptive phenomena of any volcanic focus can be made out. * * * When we descend to the shore of the Beja delle pomice by the gorge to the South East of the great obsidian flow, the slopes facing the lava are composed of immense deposits of pumice in which hundreds of holes are observable, marking the excava- tions made in search of the larger masses of this valuable rock, much of which could be seen in the numerous baskets standing at hand. The sight of the enormous THE NONMETALLIC MINERALS. 473 agglomeration of pumice and dust of a glaring white colour, cut by the action of rain and wind into fantastic shapes, stands out against the blue sky like the irregular crags, spurs and ridges of a great glacier. Along the marina are quantities of pebbles of pumice, either rounded by the torrents that descend from above or by the waves that lap the shore. When the wind blows from N. E. a veritable fleet of floating masses reaches the port of Lipari. The pumice that has been excavated is carried to the beach, and stored and sorted in sheds or caves cut out of the same pumice tuff, protected in front by a breakwater of big stones to prevent heavy seas reaching and washing away the produce. Pumice in commerce is classified as follows grosse( large size), correnti (medium), andpezzani (small) ; the large and middle size are subdivided into lisconi (flat) and rotondi (round) . The lisconi are filamentous and break less easily than the rotondi. They are also trimmed by the sorters. The lisconi. and rotondi are again subdivided into white, black, and uncertain, according to their colour. The price varies according to the quality from 50 to 2000 lire the ton. The common price for the assorted is 350 to 500 lire the ton. As much as 5000 tons a year are exported. The best pumice is that of Campo Bianco. It is also obtained at Perera, but it is in small quantity and was produced at the eruption of the Forgia Vecchia. It is a first class grey pumice and fetches from 600 to 750 lire the ton, and does not so easily break as the Campo Bianco. Also at Vulcano a grey pumice is found but the presence of included crystals render it useless for commercial pur- poses. At Castagna a commoner pumice is obtained called Alessandrina, of which brick shaped pieces are made and used for smoothing oil-cloth. 1 According to the Engineering and Mining Journal 2 a merchantable pumice has recently been found in Miller County, Idaho, but the demands for material of this nature is likely to be lessened by the putting upon the market of a German artificial product. In 1897 some 1,700 tons of pumice were mined near Black Rock, Millard County, Utah. Ground and bolted pumice is quoted as worth from $25 to $35 a ton according to quality. 3. ROTTENSTONE. The name rottenstone has been given to the residual product from the decay of silico-aluminous limestones. Percolating carbonated waters remove the lime carbonate from these stones, leaving the insoluble residue behind in the form of a soft, friable, earthy mass of a light gray or brownish color, which forms a cheap and fairly satisfactory polisher for many metals. Specimens Nos. 54150, 54153, 67390, 67791, U.S.N.M., show the material in its natural state and ground and bolted. The chemical composition of rottenstone, as may well be imagined from what has been said regarding its method of origin, is quite variable, though alumina is always the predominating constituent. Analyses show: Alumina, 80 to 85 per cent; silica, 4 to 15 per cent; 1 The South Italian Volcanoes, by H. J. Johnston-Lavis, Naples, F. Furchheiin, 1891, pp. 67-71. 2 Volume LXIV, July 24, 1897, p. 91. 474 REPORT OF NATIONAL MUSEUM, 1899. 5 to 10 per cent of carbon, and equal amounts of iron oxides and varying small quantities of lime. The material has little commercial value. 4. MADSTONES. These need but brief notice here. The fallacy of the madstone dates well back into the dark ages and perhaps beyond, and strange as it may seem continues down to the present day. Not longer ago than Decem- ber, 1898, the Washington newspapers chronicled the sale for $450 of a madstone in Loudoun County, Virginia, and from year to year very many letters are received by the Smithsonian authorities making inquiries regarding such, or possibly offering one for sale at fabulous prices. So far as the writer is able to learn, either from literature or from personal examination, stones of this class are almost invariably of an aluminous or clayey nature, and their supposed virtue is due wholly to their avidity for moisture their capacity for absorption, which causes them to adhere to any wet surface, as the tongue or to a wound, until saturated, when they will drop away. It should not be neces- sary to state, at this late day, that their curative powers are purely imaginary. The ancient bezoar stone, used in extracting or expelling poisons, consisted of a calculus or concretion found in the intestines of the wild goat of northern India. 1 5. MOLDING SAND. For the purpose of making molds for metallic casts, a fine, homo- geneous argillaceous sand is commonly employed. The physical qualities which go to make up a molding sand consist, according to Nason, 2 of elasticity, strength, and a certain degree of fineness. It must be plastic in order to be molded around the pattern; it must have sufficient strength to stand when unsupported by the pattern, and to resist the impact of the molten metal when poured into the mold. Too much clay and iron present in the sand will cause the mold to shrink and crack under the intense heat; too little will cause it to dry and crumple, if not to entirely collapse. The peculiar virtues of molding sand, as outlined above, are ascribed to the fact that each of the sand grains is coated with a thin film of clay. The accompanying table will serve to show the varying chemical character of sands thus employed, though, according to authorities 1 W. J. Hoffman, Folk Medicine of the Pennsylvania Germans, Proceedings of the American Philosophical Society, XXVI, 1889, p. 337. 2 Forty -seventh Annual Report of the Regents State Museum of New York, 1893, p. THE NONMETALLIC MINERALS. 475 quoted by Crookes and Rohrig, 1 the "quality of the sand for molding depends less on its chemical composition than on its physical proper- ties, namely, whether the grains are round, angular, scaly, etc., and whether they are of uniform size. The adhesiveness is dependent not alone on the quantity of clay, but upon the angularity of the grains, and by a mixture of smaller and larger grains. Reinhardt states that to the naked eye, a good sand should consist of particles seem- ingly uniform in size, with a sharp feel to the touch. When strewn upon dark paper it should show no dust, and when moistened with from 10 to 20 per cent of water it must be capable of being formed into balls without becoming pulpy or being too easily crushed. Constituents. I. II. III. IV. V. VI. VII. VIII. SiOo 92 083 91.907 5.683 2.177 0.415 92.913 5.850 1.249 Traces 90.625 6.067 2.708 Traces 79.02 13.72 2.40 71 86.68 9.23 3.42 0, 90 87.6 7.7 3.6 0.% 90.25 4.10 5.51 0.23 Al0 :! 5.415 Fe 2 O . and FcO 2. 498 CaO Traces MgO K 2 O 4. 58 100. 29 100.09 99.9% 100. 182 100.012 100.000 100. 43 99. .SO Of the above No. I is from Charlottenburg, Germany; No. II, a sand employed for bronze castings in Paris foundries; No. Ill, sand from Manchester, England; No. IV, from near Strom berg; No. V, from Ilsen- burg, in the Hartz Mountains; No. VI, from Sheffield, England; No. VII, from Birmingham, England, and No. VIII. from Liineburg. The sand from Ilsenburg, the composition of which is given in column 5, is stated 2 to be prepared by mixing "common argillaceous sand, sand found in alluvial deposits, and sand from solid sandstone." In preparation the first two are carefully heated to dehydrate the clay and then mixed, equal proportions of each with the same amount of sandstone. The mixture is then ground and bolted, the product being as fine as flour and capable of receiving the most delicate impressions. According to D. H. Truesdale, 3 the four essential qualities in mold- ing sand are, in the order of their importance, (1) refractoriness, (2) porosity, (3) fineness, and (4) bond. These qualities are dependent mainly upon the varying properties of siliceous sand and clay, the refractory nature being governed by the absence of such fluxing con- stituents as calcium carbonate, the alkalies, or of iron oxides. Since in nature it is not always possible to obtain the admixture of just the right proportion, artificial mixtures are often resorted to, as mentioned 1 A Practical Treatise on Metallurgy, II, p. 626. 2 Percy's Metallurgy, 1861, p. 239. 3 The Iron Trade Review, October, 1897, p. 24. 476 REPORT OF NATIONAL MUSEUM, 1899. above. W. Ferguson gives 1 the following analyses of molding sand in actual use in his foundries: Constituents No. 1, fine sand for snap work. No. 2, medium grade for medium class of work. No. 3, coarse sand for heavy ma- chine castings. No. 4, for heavy machinery in dry-sand molds. Silica 81.50 84.86 82.92 79.81 Alumina 9.88 3.14 7.03 2.18 8.21 2.90 10.00 4.44 Combined water 3.00 1.85 2.20 1.10 2.85 1.10 2.89 1.25 Magnesia 0.65 0.98 None. 0.88 Potassium No estimate. No estimate. No estimate. No estimate. Trace. Trace. Trace. Trace. Organic matter Trace. Trace. Trace. Trace. Total 100 02 98.35 97.98 99 27 Sands containing lime or alkalies, that is those containing free calcite or feldspathic granules, are sometimes liable to fusion in the case of heavy castings. It is customary in such cases to coat the surface of the mold with graphite. Sands suitable for ordinary castings are widespread, though the finer grades are often brought considerable distances, some of those used in bronze casting in America being imported from Europe. In the United States the beds are alluvial deposits of slight thickness. Large areas occur in New York State, in counties extending from the Adirondacks to New Jersey. At date of writing a very considerable proportion of the material used in the eastern United States is dug in Selkirk, Albany County, New York. (Specimen No. 61044, U.S.N.M.) Nason states that these sands occur in beds varying from 6 inches to 3 feet or even 5 feet in thickness. They immediately underlie the surface soil and overlie coarser, well stratified sand beds more nearly allied to quicksands. In gathering the sands for market, a section of land 1 or 2 rods in width is stripped of its overlying soil down to the sand, which is then dug up and carried away. When the area thus exposed is exhausted, a like area, immediately adjoining is stripped, the soil from the second being dumped into the first excavation. By this method the field, when finally stripped of its molding sand, is ready again for cultivation. It is estimated that a bed of sand 6 inches in thickness will yield 1,000 tons an acre. The royalty paid the farmers from whose land it is taken varies from 5 to 25 cents a ton. Some 60,000 to 80,000 tons are shipped annually from Albany County alone. The Selkirk molding sand is of a yellow-brown color, showing under the microscope angular and irregular rounded particles rarely more 1 Iron Age, LX, December, 1897, p. 16. THE NONMETALLIC MINERALS. 477 than 0.25mm. in diameter, interspersed with finely pulverulent matter which can only be designated as clay. The yellow-brown color of the sand is due to the thin film of iron oxide which coats the larger gran- ules. When this film is removed by treatment with dilute hydrochloric acid, the constituent minerals are readily recognized as consisting mainly of quartz and feldspar fragments (both orthoclase and a plagioclase variety), occasional granules of magnetic iron oxide, and irregularly outlined scales of kaolin, together with dust-like material too finely comminuted for accurate determination. . Many of the larger granules are white and opaque, being presumably feldspar in transition stages toward kaolin. An occasional flake of hornblende is present. The term greensand 1 is applied to the argillaceous molding sands, in an undried state, and which is employed in its native state, new and damp. The term dry mnd is used in contradistinction, to indicate a sand that must be dried by heat before being fit for use. The dry sand is stated to be firmer and better adapted than the green for molding pipes, col- umns, shafts, and other long bodies of cylindrical form. In England good molding sands are obtained from the Lower Mot- tled Sands of the Bunter (Trias) beds and from those of the Thanet (Lower Eocene). 6. MINERAL WATERS. From a strictly scientific standpoint any water is :i mineral water, since water is itself a mineral an oxide of hydrogen. Common usage has, however, tended toward the restriction of the name to .such waters as carry in solution an appreciable quantity of other mineral matter, although the actual amounts may be extremely variable. Of the various salts held in solution, those of sodium, calcium, and iron are the more common, and more rarely, or at least in smaller amounts, occur those of potassium, lithium, magnesium, strontium, silicon, etc. The most common of the acids is carbonic, and the next probably sulphuric. Classification. The classification of mineral water is a matter at- tended with great difficulty from whatever standpoint it is approached. Such classification may be either geographic, geologic, therapeutic, or chemical, though the first two are naturally of little value, and the therapeutic, with our present knowledge, is a practical impossibility. The chemical classification is, on the whole, preferable, although even this, owing to the great variation of methods of stating results used by analytical chemists, is at present attended with some difficulty. Dr. A. C. Peale, the well-known authority on American mineral waters, has suggested the scheme given below, 2 and from his writings has been gleaned a majority of the facts here given. 1 This must not be confounded with the Greensand marl, or Glauconitic sand used for fertilizing purposes, and mentioned on page 369. 2 Annual Report of the U. S. Geological Survey, 1892-93, p. 64. 478 REPORT OF NATIONAL MUSEUM, 1899. According to their temperatures as they flow from the springs the waters are divided primarily into (A) thermal and (B) nonthermal, a thermal water being one the mean annual temperature of which is 70 F. or above. Each of these groups is again subdivided according to the character of the acids and their salts held in solution as below: Class I. Alkaline. II. fSulphated. /">i T\T A -A Muriated. Any spring of water may be characterized by the presence or absence of gas when it is designated' by one of the following terms: (1) Nongaseous (free from gas). (2) Carbonated (containing carbonic- acid gas). (3) Sulphureted (containing hydrogen sulphide). (4) Azo- tized (containing nitrogen gas). (5) Carbureted (having carbureted hydrogen). In cases where there is a combination of gases such is indicated by a combination of terms, as sulphocarbonated, etc. The classes may be further subdivided according to the predominating salt in solution, as (1) sodic, (2) lithic, (3) potassic, (4) calcic, (5) magnesic, (6) chalybeate, (7) aluminous. The alkaline waters, Class I above, include those which are charac- terized by the presence of alkaline carbonates. Generally such are characterized also by the presence of free carbonic acid. Nearly one- half the alkaline springs of the United States are calcic-alkaline, that is, carry calcium carbonate as the principal constituent. The saline waters include those in which sulphates or chlorides predominate. They are mo re numerous than are the alkaline waters. The alkali-saline class includes all waters in which there is a combination of alkaline carbonates with sulphates and chlorides; the acid class includes all those containing free acid, which is mainly carbonic, though it may be silicic, muriatic, or sulphuric. The character of the salts held in solution is the same for both ther- mal and nonthermal springs, though as a general rule the amount of salt is greatest in those which are classed as thermal. Thus at the Hot Springs of Virginia one of the springs, with a temperature of 78 F., has 18.09 grains to the gallon of solid contents, while another, with a temperature of 110 F., has 33.36 grains to the gallon. Source of mineral waters. Pure water is a universal solvent and its natural solvent power is increased through the carbonic acid which it takes up in its passage through the atmosphere, and by this same acid and other organic andinorganic acids and the alkalies which it acquires in passing through the soil and rocks. The water of all springs is THE NONMETALLIC MINERALS. 479 meteoric, that is, it is water which has fallen upon the earth from clouds, and gradually percolating downward issues again in the form of springs at lower levels. In this passage through the superficial portion of the earth's crust it dissolves the various salts, the kind and quantity being dependent upon the kind of rocks, the temperatures and pressure of the water, as well as the amount of absorbed gases it contains. Both the mineral contents and the temperature of spring waters are dependent upon the geological features of the country they occupy. As a rule springs in regions of sedimentary rocks carry a larger proportion of salts than those in regions of eruptive and metamorphic rocks. Thermal springs are as a rule limited to regions of compara- tive recent volcanic activity, or where the rocks have been disturbed, crushed, folded, and faulted, as in mountainous regions. Occasional thermal springs are met with in undisturbed areas, but such are regarded as of deep-seated origin, and to owe their temperatures to the great depths from which they are derived. Distribution. Mineral springs of some sort are to be found in each and all of the States of the American Union, though naturally the resources of the more sparsely settled States have not as yet been fully developed. For this reason the table given herewith is to a certain extent misleading: Production of mineral waters in 1899 by States and Territories. State or Territory. Alabama 4 Arkansas 5 California 38 Colorado 11 Connecticut District of Columbia Florida Georgia 6 Illinois Indiana 12 Iowa 3 Kansas Kentucky 4 Maine Maryland 11 Massachusetts 39 Michigan 21 Minnesota 4 Mississippi 6 Missouri ' 12 New Hampshire New Jersey "7 New Mexico , , 5 Springs report- ing. Gallons. 38,900 48, 602 1,464,075 642, 850 338,017 168,500 17,000 128,040 858,950 162,475 40,200 63,500 1,850,132 100,380 4,439,041 3,045,400 2, 078, 700 271,500 551,876 469,800 332,000 46,800 19,917 17,442 698,493 172, 970 50, 685 10, 275 7,2.50 24, 770 101,090 25, 255 3, 320 2,718 7,032 179, 450 13,045 230,704 368,235 54,704 48,292 262,705 190,990 171,380 7,770 480 REPORT OF NATIONAL MUSEUM, 1899. Production of mineral waters in 1899 by States and Territories Continued. State or Territory. Springs report- ing. Product. Value. New York .... 46 Gallons. 4, 454, 057 8809 056 7 103 150 >0 715 Ohio 15 2 2,494,473 45 500 171,135 9 700 Pennsylvania Rhode Island 25 4 1,542,800 195,000 340,254 15 000 5 322 564 33 450 South Dakota 2 138,645 44 073 6 346 700 55 658 Texas 15 4, 729, 950 155 047 Utah 3 7 850 1 955 Vermont 6 53 917 15 869 Virginia 39 954 689 341 769 3 54 000 7 002 West Virginia 7 32 220 18 305 Wisconsin Other Statesa 30 4 4,089,329 263, 782 701,367 75, 847 Total 479 37 021 539 5 484 694 Estimated production of springs not reporting sales 62 2, 540, 597 1 463 336 Grand total 541 39, 562, 136 6, 948, 030 a The States in which only one spring for each has made a report are included here. These States are Idaho, Louisiana, Montana, and Nebraska. Uses. The mineral waters are utilized mainly for drinking and bathing purposes, the thermal springs being naturally best suited for bathing, and the nonthermal for drinking purposes. For exhibition purposes the following waters have been selected, kind and geographic distribution being the controlling factors in mak- ing up the collection. In all cases the samples are exhibited in the original bottles as put upon the market. ALKALINE WATERS. Poland Natural Spring Water, Poland Springs, Maine. Ballardvale Lithia Spring Water, Ballardvale, Massachusetts. Londonderry Lithia Spring Water, Londonderry, New Hampshire. Otterburn Lithia Water, Amelia, Virginia. Capon Springs Mineral Water, Capon Springs, West Virginia. Jackson Lithia Spring Water, Jackson County, Missouri. Algonquin Spring Water, Prince George County, Maryland. Manitou Natural Mineral Water, Manitou, Colorado. Bock Mineral Water, Jeffress, Virginia. Massanetta Spring Water, Harrisonburg, Virginia. Bethesda Natural Mineral Water, Waukesha, Wisconsin. Clysmic Natural Mineral Water, Waukesha, Wisconsin. White Rock Lithia Water, Waukesha, Wisconsin. Idanha Natural Mineral Water, Soda Springs, Idaho. Mis&isciuoi Mineral Water, Sheldon, Vermont. THE NONMETALLIC MINERALS. 481 ALKALINE SALINE WATERS. 1. 'Sulphated. Takoma Springs Water, Takoma Park, Maryland. Fonticello Lithia Water, Chesterfield County, Virginia. Tredyffrin Lithia Water, Chester County, Pennsylvania. Chairman Natural Mineral Water, Franklin County, Pennsylvania. Harris Antidyspeptic and Tonic Water, Burkeville, Virginia. Crockett's Arsenic Lithia Water, Shawsville, Virginia. Thompson's Bromine and Arsenic Springs Water, Ashe County, North Carolina. Harris Lithia Water, Laurens County, South Carolina. Stafford Mineral Water, Jasper County, Mississippi. Bladensburg Spa Mineral Water, Bladensburg, Maryland. Healing Springs Lithia Water, Bath County, Virginia. Fairchild's Potash Sulphur Water, Garland County, Arkansas. Buffalo Lithia (Spring No. 2) Mineral Water, Buffalo Lithia Springs, Virginia. Geneva Red Cross Lithia- Spring Water, Geneva, New York. W right's Epsom Lithia Water, Mooresburg, Tennessee. Veronica Natural Mineral AVater, Santa Barbara, California. 2. Murialed. Como Lithia Water, Henrico County, Virginia. Powhatan Natural Mineral Water, Alexandria County, Virginia. Blackistone Island Mineral Water, St. Marys County, Maryland. , Columbia Natural Lithia Water, Washington City. Saratoga Natural Vichy Water, Saratoga Springs, New York. Lincoln Spring Water, Saratoga Springs, New York. The Hathorn Mineral Water, Saratoga Springs, New York. High Rock Springs Water, Saratoga Springs, New York. Congress Water, Saratoga Springs, New York. Houston Lithia Water, Houston, Virginia. SALINE WATERS. 1. Sulphated. Indian Spring Water, Sligo, Maryland. Rockhill Spring Water, Rockville, Maryland. Pluto Spring Water, French Lick Springs, Indiana. Excelsior Mineral Water, Excelsior Springs, Michigan. Greenbrier White Sulphur Water, Greenbrier County, West Virginia. Geneva Lithia Water, Geneva, New York. Blue Ridge Springs Water, Botetourt County, Virginia. 2. Muriated. Anipa Spring Water, Rome, Georgia. Deep Rock Spring Mineral Water, Oswego. New York. Blue Lick A\ T ater, Blue Lick Springs, Kentucky. ACID WATERS. 1. Sulphated. Shenandoah Alum Springs Water, Shenandoah County, Virginia. Rockbridge Alum Springs Water, Alum Springs, Virginia. Wallawhatoola Sulphated-aluminous Chalybeate Water, Millboro Springs, Virginia. NAT MUS 99 31 482 EEPOET OF NATIONAL MUSEUM, 1899. 7. ROAD-MAKING MATERIALS. Roadways subject to any considerable amount of traffic demand almost invariably some sort of stone bedding to prevent their becom- ing soft or badly cut up and rutted by wheels and hoofs of horses. Until within a comparatively few years it has been the general custom to pave the streets of cities and towns with rectangular blocks of granite, trap, or other hard rock, forming thus the well-known Belgian block and Telford pavements. Such are set in regular rows and the interspaces filled with sand and sometimes with tar or asphalt. For suburban and country roads a pavement of broken stone, the invention of a Mr. L. Macadam about 1820, and known by his name, is at pres- ent the most extensivel} T used. The invention is based upon the prop- erty possessed by freshly broken stone of becoming compacted and to a certain degree even cemented when subject to heavy rolling and the impact of wheels. The finer particles, broken away by the action of the wheels, fill the interstices of the larger, and gradually bring about an induration forming a roadbed hard, smooth, and durable. Not all materials are equa% good for macadamizing purposes. If the rock is too hard ordinary travel is not sufficient to produce the desired amount of fine material, and satisfactory cementation does not ensue. If too soft it grinds away too rapidly. If the material is . decomposed, it is stated, it does not become sufficiently indurated refuses to set, as it were. Obviously the bulk matter of any roadbed must be built up of materials from near-by sources, owing to cost of transportation. For surfacing, however, materials are often carried long distances. For this purpose a hard, dense rock, such as the finer grades of trappean rocks, are now most generally used. Macadam is laid with or without a foundation of larger stones. When such is used a thickness of from 6 to 12 inches is recommended and over this is laid from 4 to 6 inches of the broken stone or "metal." Taking all points into consideration, it is probable that the best size for macadam, for hard and tough stones, such as basalt, close-grained granite, syenite, gneiss, and the hardest of primary crystallized rocks, is from 1 to 1J inches cube, according to their respective toughness and hardness, while stones of medium quality ought to be broken to gauge of from 1 to 2\- inches, and the softer kinds of stone might vary between the limits of 2 and 1\ or 2J inches, but the latter is a size which should seldom be specified. On roads for light driving it is customary to place a final surfacing of smaller stone, such as will pass a 1-inch mesh. Considerable importance is attached to the manner in which the macadam is pre- pared for use. Machine-broken stone is not considered of the same value as that broken by hand. The stones are not so regular a size and shape, and there is a greater proportion of inferior stuff. A mechanical crusher is apt to stun the mate- rial, and does not leave the edges so sharp for binding as they are when the stone is broken with a small hammer. 1 1 Circular No. 12, TJ. S. Department of Agriculture, Office of Road Inquiry, 1896. THE NONMETALLIC MINERALS. 483 The cost of macadamized roads from necessity varies ajinost indefi- nitely. The primary factors are (1) cost of labor, (2) accessibility of materials, and (3) character of country. From $2,000 to $2,500 a mile is perhaps an average figure for localities where materials are available close at hand. The collections are intended to show only the average sizes employed and the varying nature of materials. UCLA Geology-f^oDhysics Library 405 f venue Los An^.oo ,aiif. 90024 &^1 ^^^^v f Wl t >> K r* t ^ / V , fF 'jWgj iW>'f?r fir-iv 1 fc. W > '4 si 5 R L F SEE SPINE FOR BARCODE NUMBER