LI RARY UNIVERSITY Q CALIFOtNWk EARTH SCIENCff LIBRARY 2^ ,1* o METEORITES THEIR STRUCTURE, COMPOSITION, AND TERRESTRIAL RELATIONS BY OLIVER CUMMINGS JPARRINGTON, Ph.D. CURATOR OF GEOLOGY FIELD MUSEUM OF NATURAL HISTORY CHICAGO, U. S. A. 1915 PUBLISHED BY THE AUTHOR COPYRIGHT, 1915 O. C. FARRINGTON R. R. DONNELLEY & SONS COMPANY CHICAGO BAtTH PREFACE Three reasons may be assigned for ascribing peculiar interest to the study of meteorites: First. They are our only tangible sources of knowledge regarding the universe beyond us. Second. They are portions of extra-terrestrial bodies. Third. They are a part of the economy of Nature. No survey of Nature can be considered complete which does not include an account of them. For these and other reasons, the writer has long experi- enced a fascination and delight in the study of these bodies. In seeking works for his guidance, however, he has found a lamentable lack of any which treated the subject comprehen- sively. While some phases of the subject and the character- istics of many individual falls have been investigated with admirable thoroughness, the subject as a whole has not received extensive treatment. The admirable Meteoriten- kunde of Cohen would have left little to be desired had its author been permitted to carry out his broadly conceived plan, but this privilege was unfortunately denied him. Meunier's Meteorites has not been revised in recent years and Fletcher's Introduction, while a model of its kind, is limited in its scope. That the present writer has been greatly assisted by the above works and many others in the preparation of this one needs hardly to be stated. Detailed references to these works, however, were deemed to be impracticable except where it was thought that a fuller treatment of certain subjects might be desired by some readers. In such cases references have been given. Much assistance in the preparation of illustrations for this work was given the writer by the late Prof. Henry A. Ward. Mr. D. M. Barringer generously furnished photo- graphs of Meteor Crater, Arizona, and the writer is indebted to the Journal of Geology through its editor, Prof. T. C. Chamberlin, for the loan of several cuts. S31086 TABLE OF CONTENTS CHAPTER Page I. GENERAL CHARACTERS AND NOMENCLATURE . i II. PHENOMENA OF FALL j III. GEOGRAPHICAL DISTRIBUTION OF METEORITJES 34 IV. TIMES OF FALL ^ V. SHOWERS 46 VI. SIZE OF METEORITES 54 VII. FORMS OF METEORITES ... 60 VIII. CRUST .... 78 IX. VEINS .' . 85 X. STRUCTURE OF METEORITES 92 a. IRONS 92 b. IRON-STONES . . . ... . . IQI c. STONES . IO 2 XI. COMPOSITION OF METEORITES 113 XII. CLASSIFICATION j<^ XIII. ORIGIN ' . . . 205 XIV. TERRESTRIAL RELATIONS 214 XV. METEORITE COLLECTIONS . . 220 Vll LIST OF ILLUSTRATIONS PAGE THE BACUBIRITO, MEXICO, METEORITE Frontispiece Fig. i. FALL OF THE TABORY METEORITE 8 " 2. FALL OF THE AGRAM METEORITE n 3. FALL OF THE KNYAHINYA METEORITE 15 4. HOLE MADE BY THE ST. MICHEL METEORITE 22 5. METEOR CRATER, ARIZONA 24 6. CRATERS OF THE MOON ... 26 7. OLD DRAWING PERHAPS REPRESENTING A FALL OF METEORITES 28 8. DIAGRAM SHOWING EFFECT OF OBSERVER'S POSITION ON APPARENT PATHS OF METEORS 32 9. CURVE OF METEORITE FALLS BY MONTHS 39 10. CURVE OF METEORITE FALLS BY HOURS ... ... 42 11. DIAGRAM SHOWING RELATION OF TIME OF DAY TO VELOCITIES OF METEORITES 43 12. DISTRIBUTION OF INDIVIDUALS OF THE HOMESTEAD METEORITE SHOWER 47 13. INDIVIDUALS OF THE ORGUEIL, FRANCE, METEORITE SHOWER . 51 14. THE CAPE YORK, GREENLAND METEORITE 54 15. THE WILLAMETTE, OREGON, METEORITE 56 16. EL MORITO, MEXICO, METEORITE 57 17. DIAGRAM SHOWING DEVELOPMENT OF CONICAL FORM . . . 61 18. FRONT AND REAR SIDES OF CABIN CREEK METEORITE .... 62 19. FRONT SIDE OF GOALPARA METEORITE 63 ' 20. THE JONZAC METEORITE . 64 ' 21. SIDE VIEW OF WILLAMETTE METEORITE ....... 65 ' 22. THE LONG ISLAND METEORITE 66 1 23. SIDE AND FRONT VIEWS OF THE BATH FURNACE METEORITE . 67 ' 24. FRONT AND SIDE VIEWS OF THE ALGOMA METEORITE ... 69 ' 25. CHARLOTTE AND BOOGALDI METEORITES 71 1 26. FRONT END OF THE BOOGALDI METEORITE 72 1 27. ETCHED SECTION OF THE BOOGALDI METEORITE 73 1 28. THE BABE'S MILL METEORITE . 74 ' 29. THE TUCSON METEORITE 75 30. THE HEX RIVER AND KOKSTAD METEORITES 76 31. CRUST OF THE CHARLOTTE METEORITE 79 32. SURFACE OF THE JUNCAL METEORITE 80 33. MICROSCOPIC SECTION OF CRUST OF Mocs METEORITE ... 82 34. ENLARGED VIEW OF VEIN OF Mocs METEORITE 86 35. SLICKENSIDED SURFACE OF LONG ISLAND METEORITE ... 90 36. ETCHING FIGURES OF THE RED RIVER METEORITE .... 93 37. ETCHING FIGURES PARALLEL TO AN OCTAHEDRAL FACE ... 94 38. ETCHING FIGURES PARALLEL TO A CUBIC FACE 94 39. ETCHING FIGURES PARALLEL TO A DODECAHEDRAL FACE . . 94 40. ETCHING FIGURES PARALLEL TO AN ASYMMETRICAL FACE . . 94 41. TESSELATED OCTAHEDRAL FIGURES 96 42. DIAGRAM OF ETCHING FIGURES OF A CUBIC METEORITE . . 97 43. ETCHING FIGURES OF A TOLUCA METEORITE BEFORE AND AFTER HEATING 99 44. CHONDRI OF THE HOMESTEAD METEORITE AS SEEN UNDER THE MICROSCOPE- 104 ix x LIST OF ILLUSTRATIONS Fig. 45. LARGE CHONDRUS ENCLOSING A SMALL ONE AS SEEN UNDER THE MICROSCOPE 106 ' 46 MICROSCOPIC SECTION OF THE MEZO MADARAS METEORITE . 107 ' 47. DRAWING OF PYRRHOTITE CRYSTAL FROM THE JUVINAS METEORITE 142 ' 48. REICHENBACH LAMELLAE 143 1 49. PERFORATED CANYON DIABLO METEORITE 144 50. BREZINA'S LAMELLAE 150 51. DRAWINGS OF ANORTHITE CRYSTALS FROM THE JUVINAS METEORITE 167 52. DRAWING OF AN ENSTATITE CRYSTAL FROM THE STEINBACH METEORITE 172 53. DRAWING OF AN ENSTATITE CRYSTAL FROM THE STEIN RACK METEORITE 172 54. DRAWING OF A DIOPSIDE? CRYSTAL FROM THE JUVINAS METEORITE 179 55. DRAWING OF AN AUGITE CRYSTAL FROM THE JUVINAS METEORITE 180 56. FORMS OF CHRYSOLITE FROM THE PALLAS METEORITE ... 182 57. COMMON FORMS OF METEORITIC CHRYSOLITE 183 58. TYPICAL ARRANGEMENTS OF CHRYSOLITE LAMELLAE .... 184 59. EFFECT OF EARTH'S GRAVITATION ON BODIES OF DIFFERENT VELOCITIES 207 ' 60. DANIEL'S COMET 209 ' 61. A SHOOTING STAR TRAIL SHOWING INCREASE IN BRIGHTNESS . 210 ' 62. THE PLANET SATURN AND ITS RINGS 212 ' 63. MOVING THE CAPE YORK METEORITE IN NEW YORK CITY . .221 ' 64. TOWN HALL AT ELBOGEN, BOHEMIA, IN WHICH A METEORITE HAS HUNG FIVE CENTURIES 222 ' 65. METEORITE COLLECTION OF THE FIELD MUSEUM OF NATURAL HISTORY 224 METEORITES CHAPTER I GENERAL CHARACTERS AND NOMENCLATURE Meteorites are solid bodies which come to the earth from space. Their dimensions range from microscopic to many cubic feet. Their fall to the earth is usually marked by peculiar phenomena of sound and light. The masses observed to fall are for the most part of a stony nature, granular, of a grayish color, and covered with a thin, black crust. As a rule, they contain particles of metal scattered through their substance. In the material of some falls the metal more largely predominates and still others are made up wholly of metal. The nature of this metal is essentially similar in all meteorites, being iron alloyed with from five to twenty-five per cent of nickel. According to their prevailing substance meteorites may be divided into the two classes of stone and iron meteorites. It is convenient also at times to distinguish an intermediate class which may be designated as iron-stone meteorites. Most stone meteorites, as has been said, have the appear- ance of a grayish mass covered with a black, more or less shining crust. Occasionally 'the mass of the stone may be so dark as to be practically black or it may be brownish. Again it may be nearly white. Further, the crust does not always differ in color from the interior, especially in the case of brown or black meteorites. Scattered metallic grains usually characterize the substance of stone meteorites. The coherence of the stone meteorites is usually such that they do not break easily under the blow of a hammer and they take a fair polish. Some can, however, be crumbled in the fingers. The iron-stone meteorites differ chiefly from the stone 2 ; : ' METEORITES m^teorkcs in their abundance in metal. Instead of occur- ring as minute, scattered grains forming but a small per- centage of the mass of the meteorite, the metal makes up about half the mass and is often continuous. Single nodules of the metal may reach a diameter of one inch or more. Further, the metal may be so abundant as to form a matrix of a sponge-like character in the pores of which silicates are held. Thus by gradation the iron-stone meteorites pass to meteorites made up entirely of metal or iron meteorites. The metal of the iron meteorites is, when fresh, of a silver-white to grayish-white color and usually malleable. It is made up, as has been said, chiefly of iron alloyed with from five to twenty-five per cent of nickel. When found immediately upon falling, meteorites of this composition usually exhibit a blackish or bluish crust through which the silvery appearing interior gleams here and there, but any continued exposure usually causes the entire surface of such meteorites to become of a rusty-brown color. No single criterion can be given for distinguishing mete- orites from masses of terrestrial origin. Only by combining several features can the positive determination of a meteorite be made. A pitted and fused surface is an important char- acter of meteorites, yet on the one hand this may not be present and the body be meteoric, and on the other hand very similar pittings, though not often produced by fusion, may be observed on terrestrial rocks. The presence of metallic grains is generally a distinctive feature of stone meteorites, but these grains are lacking in some meteorites, and a somewhat similar appearance through the presence of scattered grains of pyrite or other mineral of metallic luster, may be seen in terrestrial rocks. The true chondritic structure when observed under the microscope may be considered a decisive mark of a meteorite, yet a few mete- orites do not have this structure. So far as the iron meteorites are concerned the presence of nickel is essential. No iron meteorites are known without nickel. Yet this alone does not prove meteoric origin since terrestrial nickel-irons are known. Terrestrial nickel-irons, however, have a percentage of nickel either lower (3 per GENERAL CHARACTERS AND NOMENCLATURE 3 cent) or higher (35 percent) than that of meteorites, so that a percentage of nickel between 8 per cent and 20 per cent is a pretty sure indication of meteoric origin. The exhibition of octahedral figures on etching is a character confined to iron meteorites, yet a metallic mass may be a meteorite and not give these figures. A pitted surface is characteristic of iron as well as of stone meteorites, but this may be destroyed by weathering; so that its absence is not a sure indication of terrestrial origin. To recapitulate: Stone meteorites usually show a pitted and fused surface, differing in color from the interior and the interior usually contains scattered metallic grains. Iron meteorites always contain nickel, usually ex- hibit a pitted surface, and frequently show octahedral bands on an etched surface. A convenient test for nickel in a mass whose meteoric origin is suspected may be made by dissolv- ing a fragment of the substance to be tested in nitric acid, adding ammonia till the acid is neutralized, boiling, add- ing a little more ammonia to make sure that all the iron has been precipitated and filtering off this precipitate (ferric hydroxide). If the filtrate shows a bluish tinge the presence of nickel is indicated, but as small amounts of nickel might not be indicated in this way a few drops of yellow ammonium sulphide should be added to the cold, clear filtrate. If nickel is present, a black precipitate of nickel sulphide will be ob- tained. In order to test this further, the liquid should be filtered and the precipitate tested with a borax bead. If nickel is present a violet bead will be obtained in the oxidiz- ing flame, changing to reddish brown on cooling. Another test for nickel in the presence of iron consists of dissolving the substance to be investigated in hydrochloric acid, boiling for a moment with a few drops of nitric acid to oxidize the iron, adding a little citric or tartaric acid to prevent pre- cipitation of the iron, neutralizing the solution with am- monia and adding a few drops of a solution of di-methyl gloxine in alcohol. If nickel is present a blood-red color will be given the solution; if not, no change of color will occur. Owing to the characters above described actual observa- tion of the fall of a meteorite is no longer necessary in order 4 METEORITES to establish its meteoric origin. In fact, the internal characters of a meteorite furnish at the present time much more reliable evidence of its origin than does as a rule the testimony of human witnesses. Of about seven hundred meteorites now recognized, only about one-half were actually seen to fall. It has been more or less customary to designate meteorites seen to fall as "falls," while those determined from internal characters were called "finds." Since all true meteorites fell at some time, however, the distinction seems superfluous. In the present work all meteorites are referred to as falls, the distinctions of separate falls being based on separate occurrence in time or place, or both. As betwesn stone and iron meteorites, it may be remarked that a far larger number of stone than iron meteorites has been observed to fall. Of about 350 observed falls only 10 have been of iron meteorites. On the other hand, among meteorite "finds," the iron meteorites largely predominate. This is chiefly for the reason, doubtless, that the iron meteorites by their relatively great weight, metallic resonance and internal silvery appearance attract the attention of the ordinary observer much more quickly than the stone meteorites. The latter show to the casual ob- server no striking differences from terrestrial rocks, and are thus usually overlooked. In order to facilitate the comparison, collection, and study of different falls it was long ago found desirable to give each fall a name. Various methods of choosing such names have been adopted: First, and most commonly, that of the province or region where the fall occurred has been used. Second, the name of the discoverer has been applied. Third, the meteorite has been named from some peculiarity of its shape or size. Illustrations of the first class are the names Mexico, Colorado, Texas, which have been given to meteoric falls; of the second class, Gibbs, Lea, Pallas, Humboldt; and of the third class, Signet, Moon, Woman. Of these methods of naming the fall, that of employing the name of the place has been found most satisfactory and has come to be generally adopted, but even this method has passed through some modifications. When but few meteorites were GENERAL CHARACTERS AND NOMENCLATURE 5 known it was sufficient to designate one as Mexico, another as Colorado, etc., but after two or more meteorites came to be known from the same state or country this method of nomenclature was obviously inadequate. A modification of this method which gained some adoption was that of giving successive meteorites from the same state or province serial numbers, such as Colorado I, 2, 3, etc., but this was long since abandoned. For the most part, names of persons or descriptive names that have been applied to meteorites are now also changed to names which show localities. Thus the Pallas meteorite is now known as Krasnojarsk, the Ainsa meteorite as Tucson, the Lea meteorite as Cleveland, etc. Some authorities, such as Berwerth* have urged that the law of priority should govern the naming of meteorites as it does that of some other objects. This would require, however, giving a number of meteorites the same name and destroy in many cases the very great advantage arising from having the name of a meteorite express the locality of its fall. Accordingly at the present time in the naming of a meteorite the plan is almost universally followed of desig- nating it by the name of the town or locality of prominence nearest to which it fell. Thus the name Castine, for in- stance, is used to designate the meteorite which fell at Castine, Maine, May 20, 1848. For the science which has for its field the study of mete- orites, several names have been proposed, but none have as yet received general adoption. Shepard suggested "astrolithology," meaning lithology of the stars, but meteorites are now considered to be quite distinct from the stars. Maskelyne proposed "aerolitics, " but the term refers to but a single group of meteorites according to a distinction made by Maskelyne himself. By using the term "meteoritics" the objection mentioned may be overcome and this name may in time gain adoption. The science of meteoritics obviously looks to many other sciences for its * Verzeichnis der Meteoriten, 1902, Ann. d. K. K. Naturhist. Hof Mus., Wien, Bd. xviii, pp. 2-3. 6 METEORITES data and growth. The science of astronomy throws light on the relations of meteorites to the other heavenly bodies; the sciences of geology, petrology, and mineralogy elucidate the relations of meteorites to the earth and its rocks and minerals, and the sciences of physics and chemistry afford means for the analysis of the structure and composition, spectroscopic characters, etc., which distinguish meteorites from other bodies. CHAPTER II PHENOMENA OF FALL The fall of a meteorite is usually accompanied, as has already been noted, by phenomena of light and sound. These phenomena may be of a startling and violent char- acter or scarcely perceptible. Their nature and extent obviously vary with the distance of the observer from the place of passage of the meteor or from its place of fall, and with the time of fall. Occasionally the passage of a meteor producing meteorites may be observed over an area of thousands of square miles. Falls occurring during the daytime may present no visible phenomena of light and occasionally no sound may be heard, but usually light or sound is observed. Brief descriptions of the phenomena which have accompanied the fall of meteorites at different periods and over various parts of the earth's surface are given following. At the fall of Tabory, Perm, Russia (Fig. i), which took place at 12:30 p. M., August 30, 1847, a fiery mass appeared in a clear sky and moved in an almost horizontal direction toward the northeast. It spread sparks in its way which left a bright, smoky trace after them, and some observers saw an illuminated stripe remaining after the mass had passed. The fiery mass remained in view only two or three seconds. Two or three minutes later, sounds like the firing of many cannon were heard. In several villages of the region, black, warm stones weighing from two to twenty pounds fell to the earth. At the fall of Mocs, Hungary, which occurred February 3, 1882, at 3 145 P.M., an intensely brilliant meteor was seen in a cloudless sky, then a rolling noise and violent detona- tions were heard. At the spot where the light was first observed, a white, cirrus-like cloud extended in the form of a white stripe from west to east. About a thousand stones 7 8 METEORITES fell at this time, the largest of which weighed 70 kilos (154 pounds). The fall which tocrk place at Sokobanja, Servia, about 2:00 P.M., October 13, 1877, was introduced by two ex- plosions like salvos of artillery, accompanied by a brilliant display of light such as attends the bursting of shells. A dense, black smoke was observed at a considerable altitude, F IG j p a ll O f the Tabory, Russia, meteorite, 12:30 P. M., August 30, 1847. and this broke up into three columns which gradually changed to a white smoke. The noise lasted for some time and resembled the firing of musketry. Soon after the first sound a number of meteorites fell over an area a mile and a half in length and a half-mile in breadth. The largest of these weighed 38 kilos (84 pounds). At Sauguis, France, at 2:30 A. M., September 7, 1868, a meteor emitting a pale green light traversed the sky and broke up, leaving a faint, whitish cloud which lasted for some time. The disappearance of this cloud was succeeded by a noise as of thunder, followed by three or four loud PHENOMENA OF FALL 9 detonations, which were heard over a wide area. The inhabitants of Sauguis heard in addition to these noises a sound like that produced by quenching hot iron in water and a dull thud. A" stone weighing about 4 pounds was found to have fallen in the bed of a small stream where it was broken to fragments. The fall of Khairpur, India, which took place September 23, 1873, at 5:00 A. M., was introduced by the appearance of a cluster of meteors in the west. Each member of the cluster is described as having exceeded in brightness a star of the first magnitude and the meteors left behind them a train from 3 to 5 in breadth. The first thought of one observer was that he was gazing at a rocket, but this opinion was soon dispelled as the object rapidly increased in bright- ness and came toward him leaving a train behind. The motion was not rapid but steady and by the time the mass had come to within about 10 of the meridian, which it passed south of the zenith, it assumed an exceedingly bril- liant appearance, the larger fragments, glowing with intense white light with perhaps a shade of green, taking the lead in the cluster, surrounded and followed by a great number of smaller ones, each drawing a train after it which, blending together, formed a broad belt of a brilliant, fiery red color. This light illuminated the whole country like an electric light. The meteor proceeded in this way till it reached a point nearly due east, paling again as it drew near the hori- zon and about 20 above the horizon appeared to go out. The train continued very bright for some time and was distinctly traceable for three-quarters of an hour after. At first it changed to a dull red, then, as the morning broke, to a line of silvery gray clouds that divided into several portions and floated away on the wind. After the disappearance of the meteor and while the train still attracted attention, there was an interval of perfect silence, then a loud report, followed by a long reverberation that gradually died away as a roll of distant thunder. A number of stones fell from this meteor, over an area 16 miles long by 3 miles wide. The largest stone found weighed about 10 pounds. The fall which took place at Orvinio, Italy, August 31, 10 METEORITES 1872, was ushered in by the appearance of what seemed to be a large star of a red color traversing the sky in a northerly direction. This increased in brilliance as it drew on and left a white train in its wake. At a certain point it became brilliantly white and vanished, leaving a luminous cloud which continued to be visible for a quarter of an hour. After the lapse of two or three minutes two reports were heard, the first like that of a cannon, the second like a series of from three to six guns fired in rapid succession. A stone weighing about 7 pounds fell at Orvinio and some fragments were thought to have been carried further northward. The fall at Hessle, Sweden, which occurred January I, 1869, at 12:20 P.M., was accompanied by a sound resembling heavy peals of thunder, followed by a rattling noise as of wagons at a gallop and ending with a sound at first like an organ tone and later like that of hissing. Many small stones fell in this shower. One struck ice close to where a man was fishing and rebounded. He picked it up and found it warm. Witnesses of the fall of the meteorite of Hraschina (Agram) in Croatia, May 26, 1751, state that about 6:00 p. M., as the sun was going down, a fiery ball appeared in the sky, which after dividing into two parts with a report like the sound of artillery, scattered more so that it appeared like a fiery chain falling from heaven. After it trailed a dark smoke which exhibited different colors. Two iron masses, one weighing 80 pounds and the other 16 pounds, were found to have fallen in a field. The accompanying view (Fig. 2) drawn by Haidinger from the accounts of witnesses represents the large and small masses, and A the cloud from which they came. At Lance, France, a fall which occurred July 23, 1872 at, 5:20 P. M., was first observed as a sudden increase of light during full sunshine. A brilliant double meteor of a rose- orange color was then seen traversing the heavens with enormous velocity toward the northeast. It separated into two luminous globes, which are said to have had the appear- ance of two candle flames proceeding horizontally. These passed out of sight at a very low elevation, their disappear- PHENOMENA OF FALL 11 ance being followed by a sharp sound without echo. The inhabitants of villages to the north saw a small cloud of smoke and heard a tremendous explosion so severe that it caused houses to shake. A large stone weighing 103 pounds which penetrated the soil to a depth of four feet was found at Lance and a smaller one about six miles distant. The trajectory of this meteorite seems to have been remarkably flat. Its velocity was calculated to have been 2200 feet per second. FIG. 2. Fall of the Agram, Croatia, meteorite, about 6:00 P. M., May 26, 1751. At the left the sun is represented as shining. The meteor which preceded the fall which took place at Weston, Connecticut, at 6:30 A. M., December 26, 1807, was first seen as a globe of fire about one-half the diameter -of the moon, rising into the sky from the north. The progress of the meteor is described as not so rapid as that of ordinary meteors or shooting stars. As the morning was cloudy the meteor passed in its course behind the clouds at intervals. The dark clouds nearly obscured it but it shone through the thinner clouds and in the clear sky it flashed with a vivid light like that of heat lightning. In the clear sky a brisk scintillation was also observed about the body of the meteor, like that of a burning fire-brand carried against the wind. A conical train of light was also seen to 12 METEORITES attend it, waving, and in length about 10 or 12 diameters of the body. The meteor disappeared about 15 short of the zenith. It did not vanish instantaneously but grew fainter and fainter "as a red-hot cannon ball would do if rapidly cooled." The whole period between the first appearance and the total extinction was estimated as about 30 seconds. About 30 or 40 seconds after this disappearance, three loud and distinct reports like those of a small cannon near at hand were heard. Then followed a rapid succession of duller reports, running into each other and producing a continued rumbling like that of a cannon ball rolling over a floor with a varying intensity of sound. Some observers compared the sound to that of a wagon running rapidly down a long and stony hill, and others to a volley of musketry protracted into what is called a running fire. This sound died away in the direction from which the meteor came. Stones fell from this meteor at three different places in the line of movement over an area about 10 miles long. The largest and last to fall weighed about 200 pounds. Especial interest attaches to the circumstances of this fall on account of the fact that the possibility of such an occurrence was at that time scarcely believed and the general opinion was expressed by the President of the United States, Thomas Jefferson, in the remark that it was easier to believe that Yankee professors would lie than to believe that stones would fall from heaven. Subsequent evidence has, however, left no doubt that the Yankee professors (Profs. Silliman and Kingsley of Yale), as well as other historians of the fall, were describing a real occurrence. The meteorite which fell at Warrenton, Missouri, about sunrise January 3, 1877, first indicated its coming by a sound described by some observers as like the whistle of a locomotive and by others as like the passage of a cannon ball through the air. To four observers the sound became louder and louder and a stone struck a tree near them, breaking off the limbs and coming to the ground with a crash. The snow was melted and the frozen ground thawed near where the stone fell, but the pieces though warm were easily handled. The stone was broken by the fall but PHENOMENA OF FALL 13 the fragments aggregated nearly 100 pounds in weight. No explosion was heard nor were any luminous phenomena noted. The meteoritic shower which occurred near New Concord, Ohio, about 12:30 p. M., May i, 1860, was introduced by a strange and terrible report in the heavens, which shook the houses for many miles about. The first report was immediately overhead and after an interval of a few seconds was followed by similar reports with such increasing rapidity that after reaching the number of twenty-two they were no longer distinct but became continuous and died away like distant thunder. Three men working in a field heard, after the first terrible report, a buzzing noise over- head and soon observed a large body descend and strike the earth at a distance of about one hundred yards. This body proved to be a large stone which buried itself about two feet beneath the surface and when obtained was quite warm. The day was cool and the sky covered at the time with light clouds. At Cambridge, Ohio, eight miles west, three or four distinct explosions were heard like the firing of heavy cannon, with an interval of a second or two be- tween each report. This was followed by sounds like the firing of musketry in quick succession which ended with a rumbling noise like distant thunder. The Cabin Creek, Arkansas, meteorite, one of the few irons and the largest iron ever seen to fall, fell at 3 :oo p. M., March 27, 1886. It gave the first indication of its approach to the party who was nearest it, a lady in a house 75 yards away, by a very loud report which caused "the dishes in the closet to rattle and was louder than thunder." Running out of the house the lady saw limbs falling from the top of a tall pine tree, 107 feet high. Three hours later a hole was found near the tree in which an iron meteorite had buried itself to the depth of three feet. The ground was warm and the iron as hot as men could well handle. The loud report which startled the first mentioned observer was heard as far as 75 miles away and was there followed by a hissing sound as if metal had come in contact with water. No luminous phenomena were reported. 14 METEORITES The fall of the iron meteorite of Mazapil, Mexico, which occurred about 9 P. M., November 27, 1885, was indicated to the nearest observer by a loud, sizzling noise as though something red-hot was being plunged into cold water, and almost instantly there followed a somewhat loud thud. The air was at once filled with a phosphorescent light with small luminous sparks suspended in it. Horses in the vicinity were much frightened. The luminous air soon disappeared and there remained on the ground a light such as is made when a match is rubbed. After the observers had recovered from their surprise they saw a hole in the ground and in it a ball of light. They feared this ball would explode and retired for a time, but returning found in the hole what looked like a stone which was too hot to handle. This the next day they found to be an iron weighing about 10 pounds. The hole was about one foot deep. At the fall of the iron meteorite of Braunau, Bohemia, which took place July 14, 1847, at 3:45 A. M., the people of Braunau were wakened from sleep by two violent sounds like cannon shots followed by a whistling and rushing sound which lasted several minutes. Those who hastened into the open air saw to the northwest in a sky in which some stars were yet visible, a small, black 'cloud. This cloud glowed and emitted tongues of light, two of which flashed to the earth. About the fiery cloud was seen one of ash- gray color which finally disappeared in the direction in which the wind was blowing. An iron meteorite weighing 48 pounds was found in a hole three feet deep, and this six hours after the fall was so hot as to burn the hands of those who touched it. About a mile away to the southeast a mass weighing 35 pounds fell through the roof of a house and near a bed where three children were sleeping. At the fall of the Rowton, England, meteorite, which took place at 3:40 P. M., April 20, 1876, a strange, rumbling noise was heard, followed almost instantaneously by a startling explosion resembling a discharge of heavy artillery. About an hour later a hole was found in the ground and at a depth of 1 8 inches in this hole there reposed an iron meteorite weighing 7^ pounds. PHENOMENA OF FALL 15 At Quenggouk, India, a meteor burst into view at about half-past three A. M., December 27, 1857. It was in the western quarter of the sky. It sped across the sky in an almost due easterly direction seeming "three times as large as the full moon" and with a blinding brilliancy of light. Far behind its brilliant forward point there trailed a great, FIG. 3. Fall of the Knyahinya, Hungary, meteorite, 5.00 p. M., January 9, 1866. luminous, variegated nebulous cloud. A terrific explosion was heard, followed by lesser ones and a protracted rumbling. Three small stones were found to have struck the ground. At the fall of Knyahinya, Hungary (Fig. 3), which, took place about 5:00 p. M., January 9, 1866, those nearest the point of fall heard sounds like cannon-shots followed by a noise like the boiling of water and a long roll. At the same time a cloud of smoke appeared in the sky from which stones fell. These observers saw no light, but those at a distance 16 METEORITES of 10 or 12 miles saw a fire ball of the color of white-hot iron with edges of ultramarine blue. Some saw this divide in two. About 1000 stones, ranging in size from 2 grams to 300 kilograms, were precipitated over an area about 9 miles long by 3 miles wide. The stones picked up immediately after the fall were described as being lukewarm or warm. The largest stone found penetrated the soil to a depth of II feet, entering in an oblique direction. From subsequent measurements it was calculated that the meteor first ap- peared at a height of 7^/4 miles and dropped almost directly downward. Phenomena of especial impressiveness seem to have at- tended the fall of Homestead, Iowa, which took place Febru- ary 12, 1875, about 10 p. M. A meteor was seen moving north and east and from the first the light of the meteor could hardly be tolerated by the naked eye turned full upon it. Several observers who were facing south at the first flash, .say that upon looking full at the meteor it appeared to them round, and almost motionless in the air, and as bright as the sun. Its light was not steady, but sparkled and quivered like the exaggerated twinklings of a large fixed star, with now and then a vivid flash. To these observers, all of whom stood near the meteor's line of flight, its size seemed gradually to increase, also its motion, until it reached a point almost overhead, or in a direction to the east or west of the zenith, when it seemed to start suddenly, and dart away on its course with lightning-like rapidity. The observers who stood near to the line of the meteor's flight were quite over- come with fear, as it seemed to come down upon them with a rapid increase of size and brilliancy, many of them wishing for a place of safety but not having time to seek one. In this fright animals took part, horses shying, rearing, and plunging to get away, and dogs retreating and barking with signs of fear. The meteor gave out. marked flashes in its course, one more noticeable than the rest, when it had com- pleted about two-thirds of its visible flight. All observers who stood within twelve miles of the meteor's path say that from the time they first saw it, to its end, the meteor threw down "coals" and "sparks." PHENOMENA OF FALL 17 Thin clouds of smoke or vapor followed in the track of the meteor and seemed to overtake it at times, and then were lost. These clouds or masses of smoke gave evidence of a rush of air with great velocity into the space behind the meteoric mass. The vapor would seem to burst out from the body of the meteor like puffs of steam from the funnel of a locomotive, or smoke from a cannon's mouth, and then as suddenly be drawn into the space behind it. The light of the meteor's train was principally white, edged with yel- lowish green throughout the greater part of its length, but near to the body of the meteor the light had a strong red tinge. The length of the train was variously estimated, but was, probably, about 9, or from seven to twelve miles, as seen from Iowa City. The light about the head of the meteor at the forward part of it, was a bright, deep red, with flashes of green, yellow, and other prismatic colors. The deep red blended with and shaded off into the colors of the train at the part following; but the whole head was enclosed in a pear-shaped mass of vivid white light next to the body of the meteor, and the red light fringed the white light on the edges of the figure, and blended with it on the side presented to the eye. From three to five minutes after the meteor had flashed out of sight, observers near the south end of its path heard an intensely loud and crashing explosion that seemed to come from the point in the sky where they first saw it. This deafening explosion was mingled with, and followed by, a rushing, rumbling, and crashing sound that seemed to follow up the meteor's path, and at intervals, as it rolled away northward, was varied by the sounds of distinct explosions, the volume of which was much greater than the general roar and rattle of the continuous sounds. This commotion of sounds grew fainter as it continued, until it died away in three to five explosions much fainter than the rest. From one and a half to two minutes after the dazzling, terrifying, and swiftly moving mass of light had extinguished itself in five sharp flashes, five quickly recurring reports were heard. The volume of sound was so great that the rever- 18 METEORITES berations seemed to shake the earth to its foundations, buildings quaked and rattled, and the furniture that they contained jarred about as if shaken by an earthquake; in fact, many believed that an earthquake was in progress. Quickly succeeding, and in fact blended with the explosions, came hollow bellowings, and rattling sounds, mingled with a clang, and clash, and roar that rolled slowly southward as if a tornado of fearful power was retreating upon the meteor's path. The phenomena observed in the fall of a meteorite are due chiefly to the resistance of the air, some of the effects of which may be considered in a general way as follows: A body in moving through any medium such as air or water experiences a certain resistance; for the moving body sets in motion those parts of the medium with which it is in contact, and thereby loses an equivalent amount of its own motion. This resistance increases with the surface of the moving body; thus a soap-bubble or a snowflake falls more slowly than a drop of water of the same weight. It also increases with the density of the medium; in rarefied air it is less than in air under the ordinary pressure; and in this again it is less than in water. The resistance also increases with the velocity of the mov- ing body, and for moderate velocities is proportional to the square; for, supposing the velocities of a body made twice as great, it must displace twice as much matter, and must also impart to the displaced particles twice the velocity. For high velocities the resistance in a medium increases in a more rapid ratio than that of the square, for some of the medium is carried along with the moving body, and this, by its friction against the other portions of the medium, causes a loss of velocity. Light bodies fall more slowly in air than heavy ones of the same surface, for the moving force is smaller compared with the resistance. The resistance to a falling body may ultimately equal its weight; it then moves uniformly forward with the velocity which it has acquired. Thus, a raindrop falling from a height of 3000 feet should, when near the PHENOMENA OF FALL 19 ground, have a velocity of nearly 440 feet per second, or that of a musket-shot; owing, however, to the resistance of the air, its actual velocity is probably not more than 30 feet per second. The slowing down by the resistance of the air, of a body having the velocity of a moving meteorite, has been cal- culated for a number of special cases by Schiaparelli. He found that a ball i^ inches in diameter with a specific gravity of 3.5 and having an initial velocity of 9 miles per second would have its velocity reduced to ^3 of a mile per second on arriving at the point where the barometric pressure is Vee that at the earth's surface. If the ball had an initial velocity of 40 miles per second the reduction of velocity in the early stages of its fall would be much greater, and on arriving at the point where the atmospheric pressure is Vee that at the earth's surface it would be reduced nearly to that of the ball which started with a velocity of 9 miles per second. The changes are shown in full in the following table: Initial velocity 72,000 meters Initial velocity 16,000 meters (40 miles) per second (9 miles) per second Remaining Atmospheric Remaining Atmospheric velocity, pressure velocity, pressure in meters in mm. in meters in mm. 72,000 o . 0600 16,000 o . oooo 60,000 o . 0005 14,000 o . 0064 48,000 0.0013 12,000 0.0162 36,000 0.0031 10,000 0.0322 24,000 o . 0082 8,000 o . 0620 1 2,000 o . 03 58 6,000 o.i 280 8,000 0.0816 4,000 -355 4,OOO 0.3151 2,OOO 1-2293 2,000 i . 2489 i ,000 4 . 2986 1,000 4.3182 500 11.6192 500 ii .6388 The same author has calculated that for a body to reach the earth with a velocity of 500 meters (^3 mile) per second it must, if of the specific gravity of a stone meteorite, have a diameter of 2.61 meters (8 feet), and if an iron meteorite, 1.17 meters (3 feet). Niessl* has calculated for several observed meteoric falls the height above the earth at which their initial velocity *Sitzb. Wien Akad., 1884, 89, 2, 283-293. 20 METEORITES was overcome and from which they fell under the influence of the earth's gravity alone. These heights were: Homestead 3.7 km. Krahenberg 8.2 km. Mocs 8.4 km. Weston 1 1 . 1 km. Knyahinya. . . 11.9 km. Braunau 14.8 km. Orgueil 23.0 km. Pultusk 41 . 5 km. Hraschina 46.7 km. The slowing up of a meteorite by the resistance of the air exerts a powerful disruptive force upon it, since the rear of the meteorite tends to travel with a planetary velocity while the forward part is being checked. Thus Hauser calculated that a meteorite having a volume of a cubic meter and being a square meter in section would, if moving at a velocity of 30 miles per second, develop an internal disruptive force of nearly 3 billion kilogram-meters on arriving within 16 miles of the earth's surface. That this force would tend to burst the meteorite there can be no doubt. The enormous heat developed by such a checkingof velocity or the conversion of its motion into heat, should also be con- sidered. Thus a body weighing one pound, and moving 25 miles a second, has momentum sufficient to raise (25 x 528o) 2 -i-2g = 27 1, 500,000 pounds one foot. By Joule's equivalent, the raising of 772 pounds one foot corresponds to the heat necessary to raise one pound of water one degree Fahren- heit. If the capacity of the meteoric substance for heat is 0.2 (that of iron is 0.12), the loss of a velocity of 25 miles would be equivalent to heating (27 1,500,000 -5-0.2)^-77 2 = 1,760,000 pounds of the substance one degree Fahrenheit, if the whole of the motion was transformed into heat. The sounds like thunder usually accompanying the fall of a meteorite, are doubtless due, as in the case of lightning, to the explosive shock given to the surrounding air by the sudden heating of the air in the vicinity of the passing meteorite. The pressure thus produced is, according to Thomson* in the case of lightning, equal to ten atmospheres * Science, Dec. 17, 1909. PHENOMENA OF FALL 21 and is probably not less in the fall of the average meteorite. The prolonged and varying rolling sound is also due as in the case of lightning to irregular movements of the meteorite in its course through the air. The first sound heard comes from the part of the path nearest the observer and then follows that derived back along the meteor's path. Any twistings and bendings of the course of the meteorite will cause blendings and separations of the sound waves which will give varying effects. The effect of the impact of a meteorite upon the earth depends among other factors upon the velocity of the mete- orite and the nature of the surface upon which it falls. So far as the velocity of the meteorite is concerned, all evi- dence indicates, as already noted, that the meteorite loses its planetary speed in the upper layers of the atmosphere and falls during the latter part of its course like any free falling body. Thus the velocity of the Middlesbrough meteorite on striking the ground was calculated by Herschel to have been 412 feet a second. Borgstrom reckoned the velocity of the Hvittis meteorite from the depth of the hole which it made in a stiff loam to have been 584 feet a second; that of the St. Michel meteorite (Fig. 4) from similar observations to have been between 563 and 710 feet a second, and that of the Shelburne meteorite to have been 515 feet per second. The depth to which a meteorite will penetrate obviously de- pends much upon the nature of the soil. A meteorite strik- ing upon a ledge of rock as did that of Long Island, will, if it is a stone meteorite, itself be shattered. On the other hand, when striking soil meteorites may enter it to a considerable depth. The Hvittis and St. Michel meteorites above men- tioned passed the one into stiff clay and the other into mo- rainic material to a depth of about two feet each. The largest stone of the Estherville shower, weighing 437 pounds, penetrated stiff clay to a depth of eight feet, and the next smaller stone, weighing 170 pounds, embedded itself in sim- ilar material to a depth of five feet. These were, however, stones of higher specific gravity than those previously men- tioned. The Farmington meteorite weighing 180 pounds 22 METEORITES went into hard clay to a depth of four feet. The Kilbourn meteorite, a stone about the size and shape of a man's fist, FIG. 4. Hole in ground made by fall of the St. Michel meteorite. After Borgstrom. passed in succession through a barn roof composed of three thicknesses of shingles, a hemlock board an inch thick, and another hemlock board of the same thickness about four PHENOMENA OF FALL 23 feet below this. The direction of penetration of a meteorite is not always vertical, since the direction of motion of the meteorite is sometimes tangential. Thus the largest stone of the Knyahinya fall, a stone which weighed 660 pounds, reached a depth in the earth of eleven feet, but in a direc- tion inclined about 27 from the vertical so that a stake driven down in the center of the hole which it made failed to strike it. Motion in a direction much inclined from the vertical may account for the apparent lack of forceful im- pact observed in the fall of many large iron meteorites. These masses with their weight of many tons should seem- ingly reach to great depths on striking the earth, but little evidence of such penetration has yet been secured. As a rule these masses are found near the surface. The most extensive terrestrial effect which has ever been ascribed as possibly due to meteoritic impact is to be seen at Canyon Diablo, Arizona, at the point of fall of the Canyon Diablo meteorites. The immediate locality of the fall is known as Coon Butte or Meteor Crater. Here a circular depression about 4000 feet in diameter and 570 feet deep occurs in the surface of an otherwise comparatively level plain. (Fig. 5). The walls of this "crater," as it has been called, are composed of limestone and sandstone and the layers of these rocks dip away from the center of the crater at varying angles. Along the southern wall the lime- stone and sandstone have been lifted vertically more than 100 feet. At its highest point the crater rim is 160 feet above the outlying plain, and at its lowest 120 feet. The mass of the crater rim is composed of loose, unconsolidated material varying in size from microscopic dust to blocks weighing hundreds of tons. The floor of the crater is com- paratively level but has probably been built up by inwash from the sides. Special significance is attached by investi- gators of the region to the presence of a gray or white sand- stone, much of it in the form of a rock flour which is found in the floor of the crater and about its rim. This sandstone is regarded, on account of its structure, as showing signs of metamorphism by heat. Mixed with this material are part- icles of nickel-iron. At a depth of 820 feet below the floor 24 METEORITES METEOR CRATER- COCON1NO COUNTY- ARIZONA VIEW OF GRATP. FHC 'oTAr.GE OF SEVERAL MILEi. FIG. 5. Meteor Crater, Arizona. It is about this area that the Canyon Diablo meteorites are found. PHENOMENA OF FALL 25 of the crater undisturbed strata of red and yellow sandstone are found. Scattered about over the floor of the plain iron meteorites have been found in numbers reaching thousands, and in weight aggregating several tons. Within the crater itself only a few small meteorites have been found. The peculiar topographic form and the associated meteorites lend considerable plausibility -to a hypothesis which has been urged by several observers but especially by D. M. Barringer.* This hypothesis ascribes the formation of the crater to the impact of a huge meteorite which was wholly or in part metallic. The dimensions of this meteorite were calculated by Barringer and Tilghman from the size of the crater to have been about 500 feet in diameter. Complete proof of the correctness of the hypothesis would be obtained by rinding within the crater above the un- disturbed sandstone a meteoric mass or many of them which would together approximate the size mentioned. Although a number of borings have been made in search of such a mass or masses, none has as yet been found. Search fo.r such a mass with magnetic instruments has likewise given negative results. It has been suggested that the enormous force of impact might have rent the mass of the impinging body into minute fragments, some of which are represented by the nickeliferous particles found in the white sand, and hence no large mass now exists. No certain conclusion can as yet be drawn in regard to this view. An alternative to the hypothesis of meteoritic origin is to assume that the crater originated from some terrestrial force, and that the occurrence of such meteorites as have been found here is purely coincidental. This was the conclusion of G. K. Gil- bert, who ascribed the formation of the crater to a steam explosion of volcanic origin. The opponents of this view, however, state that there are no known volcanoes or hot springs near the region to afford the heat necessary for such an action. It seems as yet therefore impossible to give final decision as to the origin of the "crater." The resemblance of this topographic feature to that of the so-called volcanoes of the moon has often been re- *Meteor Crater: Philadelphia 1909. Published by the author. 26 METEORITES FIG. 6. Craters of the moon. They resemble Meteor Crater, Arizona, in form and by some have been thought to be formed by the impact of meteorites. PHENOMENA OF FALL 27 marked and it has been suggested that the moon craters are in reality impact pits caused by the striking of large mete- orites upon the moon's surface (Fig. 6). Those who urge this view find cause for the greater effect of such impact upon the moon as compared with the earth in the absence of an atmosphere from the moon. There is thus nothing upon the moon to burn up falling bodies before they reach its surface. There can be no doubt that the earth's atmos- phere affords an immense protection to its inhabitants in burning up bodies which would otherwise reach the earth's surface but that meteorites sufficiently large to make the great craters on the moon's surface have ever fallen on that body seems somewhat questionable. Meteorites show little warmth when they arrive upon the earth. The stone meteorites at any rate are almost always spoken of as being " milk warm " or " barely warm " by those who pick them up immediately after their fall to the earth. Neither are there any indications of any heating effect where meteorites have struck the earth. No baking of the soil or charring of vegetation can be observed. Meteorites have also fallen in haystacks, within barns or in other places where a little heat might start a fire but have never pro- duced any incendiary effects so far as known. This lack of heat is contrary to the general belief, the common opin- ion being that meteorites are intensely hot when they reach the earth. This opinion is evidently based on the brilliant light emitted by meteors in their course in the atmosphere. A little consideration of the matter, however, will convince one that no heating phenomena should be expected for three reasons: 1. The substance of stone meteorites is a poor conductor of heat. 2. The period in which they might acquire heat is ex- tremely short, but a few seconds at most. 3. Any portion of their surface sufficiently heated to become in a condition even approaching viscosity is imme- diately removed by the pressure of the surrounding air. With the iron meteorites the case is somewhat different since they are much better conductors of heat than stone 28 METEORITES meteorites. They, therefore, generally possess considerable warmth when picked up immediately after their fall. The Cabin Creek meteorite is described as being "as warm as could be handled" after being dug from a hole three feet deep. The Mazapil meteorite was so warm that it could be "barely handled" on removal. The heat emitted even in these cases, however, was not great. Any accounts, therefore, of intense heat being displayed by meteorites can usually be assumed to be false, the observer's previously formed opinion probably coloring his testimony if his testimony is sincere. No meteorite fall has ever positively been known to have been destructive to human life. Accounts pur^ porting to describe such catastrophes prove on in- vestigation to have come either from times or countries so remote that they cannot be verified. Many accounts of such an occurrence come to us from earlier times, and the scene here pictured (Fig. 7) probably illustrates destruc- tion believed by the early artist to have been caused by meteoric stones falling from the skies. But no well authenticated occurrence of the sort is known. Perhaps the most narrow escape which has ever been experienced was that of three children in Braunau at the time of fall of that meteorite in 1847. This meteorite, an iron weighing nearly 40 pounds, fell in a room where these children were sleeping and covered them with debris, but they suffered no serious injury. Other meteorites have fallen near human beings but never have struck them so far as credible information goes. That personal- injury or death might be caused by the fall of a meteorite is entirely possible, in fact is likely to occur at some time. It is re- markable that some falls, such for instance as the showers FIG. 7. Old drawing perhaps repre- senting a fall of meteorites. PHENOMENA OF FALL 29 in Iowa which occurred in fairly thickly settled communities, should not have caused serious injury to the inhabitants. Injury to animals from falling stones has perhaps occurred in a few instances. Cattle were said to have been struck by falling stones in the shower of Macao, Brazil, in 1836, and a dog was reported to have been killed by a meteoric stone in Nakhla, Egypt, in 1911. The evidence in regard to these occurrences is not, however, altogether satisfactory. Buildings have several times been struck by meteorites and usually have been penetrated by them. Besides the build- ing mentioned above as struck by the Braunau meteorite, a 12-pound individual of the Pillistfer fall fell through a tile roof and a floor of a building, and the Kilbourn meteorite penetrated a barn. Buildings were also penetrated by the Aussun, Barbotan, Benares, and Massing meteorites. The following recommendations to observers on occasions of the falls of meteorites, describing the points of information most desirable to be recorded regarding their characters and appearance, were published by a committee of the British Association in 1878:* "In recording observations on the passage of a meteor across the sky, the points which it is most desirable to arrive at are: such data as will allow of our definitely noting the direction of its path and its point of extinction, the duration of the luminous phenomenon, and of individual phases of it, the apparent magnitude of the meteor, the luminosity as compared with other brilliant objects and the changes which it may itself exhibit in this respect during the transit, the duration of the train (or i streak'), and the changes it may undergo before extinction (whether it fade away simul- taneously along the entire length, or break up into a chain of luminous fragments); also, in cases where the streak is one of great persistence, the manner of its final disappear- ance; again, when the meteor has been observed near the time of sunrise, or sunset, what change it wrought in the appearance of the visible train by the increasing or waning light of the sky. The sound attending its passage, if any, and the character of the sound, as regards intensity and dura- * Rept. Brit. Assn. Adv. Sci., 1878, pp. 375-377. 30 METEORITES tion, whether single and well defined, or a series of minor explosions closely following one another should be noted and finally, the time of appearance, and that of the inter- val before the explosion is heard. "While it is hardly possible for one observer to record all the data referred to, he should not fail to note such of them as may have come clearly within his observation. Other spectators may have remarked what he may have missed, and their joint observations may enable us to arrive at a complete physical history of the meteor in question. "It is desirable to determine two points of the track of the meteor, as far asunder as possible the points of appearance and extinction are to be preferred and to indicate the former by reference to some star or constellation which it overlies, and the latter by some object on the hori- zon against which it is projected. In cases where the meteor is seen in daytime, the data to be arrived at are the points of appearance and its angular altitude. The former may be estimated by noticing what conspicuous object lies vertically below it on the horizon; a village or a mountain peak. The more distant the object is from the spectator, the more accurate will be the determination of this element of the observation. If objects to which reference can be made should be wanting, the direction may be temporarily noted, and subsequently determined by the aid of a compass- needle. To learn an angular altitude we dare not trust general conclusions, however carefully arrived at; even experienced observers may be misled in such cases. If a vertical object, say the roof of a house, or the top of a tree, happens to lie in the direction under consideration, the observer should approach it till the line of sight of the origin of the course of the meteor skirts the summit of the ter- restrial object. The observer has now to determine how far he is removed from the object selected, its vertical height above the plain on which both are situated, and the distance above the ground of his own eye, in order to be in a position to determine the angular elevation of the point of appearance of the meteor, the position of which he desires to ascertain. PHENOMENA OF FALL 31 ; 'The apparent path of the meteor is often represented by a line like a bow; in other words, the meteor apparently ascends, culminates, and then takes a downward course. This motion is, however, for the most part apparent only; and is a consequence of the varying inclination which a straight line appears to form with the horizon at different points along its course. The observer should endeavor to determine as accurately as possible the apparent inclination at those points of the meteor's arc, or line of flight, which can be most readily identified, such as the beginning and the end of the track, or those where a break in the luminous train occurs, as well as that portion which lies parallel to the horizon. The point of extinction should especially be noted, and this is the more readily accomplished from the fact that the attention has been steadily directed to observ- ing the luminous phenomena preceding it. In regard to the point of appearance, it is of importance to determine whether the impression made on the observer was that he had witnessed the blazing forth of the meteor in the sky, or whether the meteor had entered his field of vision, and a portion of its luminous track had not been seen by him. "It is, moreover, of importance to arrive at a knowledge of the length of time occupied by the meteor in traversing the sky; this may sometimes be learned by counting the ticks of a watch, or by advancing in the direction of the object at a uniform rate, and counting the paces taken during the observation. It should also be noted whether the meteor moves onward with an accelerating or diminish- ing velocity. "The brilliancy of a meteor larger than the fixed stars of different magnitudes can most conveniently be compared with the light of Venus or Jupiter; and in the case of the largest meteors, with the apparent brilliancy and magnitude of the moon in her several phases. The colour exhibited by the meteor should also be carefully observed, and any change of hue along any part of the path should be recorded. The luminous train left after the disappearance of the meteor is sometimes very persistent, and often terminates in 32 METEORITES a cloud, faintly visible. Any peculiar structure exhibited by the train, or cloud, should be sketched on paper. "The sound attending the flight of the meteor usually consists either of several distinct explosions, or a crackling, rolling detonation. The closest attention should be given after the extinction of the meteor, for the arrival of the sound and the length of the interval should be carefully noted with the watch. "Of the many points which, as has been shown, it is de- sirable that a record should be made, an individual observer 7 \ / *\ \ ' V / \ \ / ,'' ^* * I^EH 5 ^ ^ ' Atmosphere Earth FIG. 8. Diagram showing effect of position of observer on the appar- ent paths of meteors. The actual paths are AB, but the apparent paths as seen at O are AC. After Moulton. can obviously determine but a few; all those of them, how- ever, to the accuracy of which he can certify, are of value since other observers may supply the missing data, and the whole may be collected. "If a meteorite has fallen, visit the spot where it struck the ground, and examine the hole which it formed. Deter- mine the depth, and especially notice the direction of the cavity in respect to the points of the compass. Ascertain whether the meteorite was removed from the ground soon after its descent, and whether any observations had been made at the time respecting its temperature. Make a note of the material forming the surface layer, and state whether it was moist or dry. Further inquiries in the neighborhood may lead to the discovery of other meteorites which had fallen at the same time, and at points not un- frequently miles distant. They may vary greatly in size; and stones as small as a pea or bean may be sought for." PHENOMENA OF FALL 33 To these suggestions may be added in the case of a meteoritic shower the recommendation that the distribution of the meteorites with respect to their size should be ob- served and the extent and dimensions of the area over which they have fallen should be determined as far as possible. Also an effective means of measuring the height of a meteor has been suggested by Sir Robert Ball. This method re- quires that two observers situated a number of miles apart should note the meteor and its direction from them. . Then on a map of any convenient scale a straight wire should be raised from the position of each observer in the direction in which he saw the meteor. The point of intersection of the wires will show the true position of the meteor and a per- pendicular let fall from this point will show its height ex- pressed in the scale of the map. Further, the effect of the position of the observer on the apparent paths of meteors should be considered. Thus in the accompanying figure (Fig. 8), there are represented the paths of three meteors which are parallel, AB, but their apparent paths as seen by an observer at O will vary in di- rection, being the lines AC. It is by continuing the lines backward on the celestial sphere that the point from which the meteors came can be determined. CHAPTER III GEOGRAPHICAL DISTRIBUTION OF METEORITES Broadly speaking, we know no fundamental reason why meteorites should be any more numerous upon one part of the earth's surface than upon another. Compared with the vast area of space in which meteorites wander, our earth is but a point, and moreover a rotating and wabbling point, ever presenting new surfaces to the portions of space in which it is traveling. The marksman who displays his skill by shooting glass balls thrown into the air would have the difficulty of his task enormously increased if he should endeavor to strike successively the same point upon the ball, especially if it had in addition to its forward motion one of rapid rotation about a wabbling axis. Yet this is what a falling meteorite must do if it is to reach any particular point on the earth's surface. These considerations make it difficult to believe that any par- ticular portion of the earth's surface is more likely than another to receive meteorite falls. However, knowledge of falls requires observers or finders and they must be persons of sufficient intelligence to recognize the nature of their finds. Hence a map of the localities from which meteorites are known shows by far .the larger part of them in civilized countries and the falls apparently the more abun- dant the greater the population. Of 634 known meteorites, 256 have been found in Europe and 177 in the United States. Thus more than two-thirds of the known number are from less than one-eighth of the land surface. We have no reason to suppose that these regions actually receive more mete- orites than others less intelligently populated and any ap- parent excess or lack of meteorites in any given locality must be considered in the light of these facts. Taking these facts into consideration, however, there seem to be certain inequalities of distribution of meteorites which may 34 GEOGRAPHICAL DISTRIBUTION OF METEORITES 35 be of some significance. Meteorites seem to be somewhat more abundant in mountainous or elevated regions than in those of an opposite character. Thus they seem to be more abundant near the Himalayas in India, near the Alps in Europe, and about the high peaks of the southern Appala- chians in the United States. The largest iron meteorites of North America are nearly all found in the Cordilleras. Whether such regions exert a greater gravitational force or whether they present an actual physical obstacle to the pas- sageof a meteoriteis uncertain, but it is probable that if either agency is operative it is the obstructive one. Another seem- ing difference in distribution may be noted if the meteorites of the two hemispheres of the world be compared: Thus of 256 meteorites known from the western hemisphere, 182 are irons and only 74 stones; while from the eastern hemisphere, of 378 known, 299 are stones and only 79 are irons. Ber- werth has sought to account for the excess of irons in the New World by the suggestion that the dry air of the desert areas which abound in this hemisphere has preserved mete- orites fallen in long distant periods while those of a similar age in the other hemisphere have been exposed to a moist climate and have for the most part been decomposed. It is true that many of the iron meteorites known from the western hemisphere occur upon the Mexican and Chilean deserts, but quite as many come from the southern Appa- lachians, where a comparatively moist climate prevails. There are also numerous desert areas in the Old World per- haps as fully explored as those of the New, so that on the whole the above explanation seems inadequate. Other remarkable groupings of meteorites with regard to their geographical distribution may be noted when areas smaller than hemispheres are compared. Thus of a total of nine meteorites belonging to the class of howardites, five have fallen in Russia. Of the nine meteorites known be- longing to the class of carbonaceous meteorites, three have fallen in France and two in Russia. Again small areas of equal extent and equally well popu- lated vary curiously in their number of meteorite falls. Within the state of Illinois, for instance, no meteorite is 36 METEORITES . known ever to have fallen, while in the state of Iowa, which has about the same area, but a smaller population, four falls have been noted, and from the state of Kansas, which has a larger area than Illinois, but a smaller and less uni- formly distributed population, twelve meteorites are known. CHAPTER IV TIMES OF FALL Considering meteorite falls by years it should be remem- bered that previous to the nineteenth century little reliable record of such falls is available. Single falls are known for the years 1492, 1668,' 1715, 1723, 1751, 1766, 1773, 1785, 1787, 1790, 1794, 1795, and 1796, and two falls each for the years 1753, 1768, and 1798. Moreover, for the early part of the nineteenth century the record is not very complete, since during that period the possibility of meteorite falls was yet much doubted. But from 1800 to 1910, 331 falls may be accepted as well authenticated as to their month and year. During this period eleven years show no falls whatever. These years are, 1800, 1801, 1809, 1816, 1817 1832, 1839, 1888, 1906, 1908, and 1909. Of these the years of the last decade will probably have falls to their credit after a time, since the record of falls usually lags somewhat behind their occurrence. The largest number of falls shown in any year during the period is n in 1868. The years 1865, 1877, and 1886 show 7 each. All the other years show from i to 6 falls each. The full record by years beginning with 1800 is as follows: I8OO O 1817 O 1834. 2 1851 2 1801 o 1818 3 1835 3 1852 4 1802 i 1819 2 1836 3 1853 3 1863 3 1820 i 1837 i 1854 i 1804 2 1821. i 1838 5 1855 4 1805 2 1822 5 1839 o 1856 3 1806 i 1823 2 1840 3 1857 6 1807 . 2 1824 3 1841 3 1858 4 1808 3 1825 2 1842 3 1859 5 1809 o 1826 2 1843 5 1860 5 1810 2 1827 3 1844 3 1861 3 1811 2 1828 i 1845 3 1862 2 1812 4 1829 3 1846 4 1863 6 1813 2 1830 2 1847 2 1864 3 1814 2 1831 2 1848 3 1865 7 1815........ 2 1832 o 1849 i 1866 6 1816 o 1833 i 1850 2 1867 2 37 38 METEORITES 1869 . . . . . 6 1880 3 1891 2 IQO2 c 1870 i 1881 . .. . . 2 1892 -I IQO'? i 1871 1872 1873 1874 l8?q 3 ... 4 3 - 5 c 1882 1883 1884 1885 1886 ... 4 3 3 - - 4 7 1893 1894 1895 1896 1807 . . . 4 - 3 - 3 ' * 1904 1905 1906 1907 IQO8 - 3 . . . o i o 1876 c 1887 . 6 1808 1 IQOQ o 1877. . 7 1888.. o 1800. . e 1878 5 1889 5 1900 3 350 This record on the whole seems to indicate a comparatively uniform supply of meteorites, which is the more remarkable when one considers the various chances affecting the observa- tion of their fall. The record seems to afford no evidence of cycles or periodicity. As already remarked, a large allowance for unrecorded meteorites must be kept in mind. Yet that those recorded are probably typical of the whole seems to be indicated by the fact that while opportunities for observation of meteorite falls have probably continually increased since 1800, the record by decades shows the decade from 1860 to 1870 to considerably exceed in number of falls either of the two succeeding ones. Passing from the falls by years, the falls by months may be examined. Such an examination should have an especial significance in showing the relations which meteorites may have to well-known star showers. Two of the best known of these showers occur in August and November. If meteorites are related to these, these months should show a larger fall than others. If meteorites are not related to these, no special increase for these months should be shown. On compiling the results it is found that the months of May and June exhibit the greatest number of falls. The number for November falls below the average and that for August rises only slightly above. The evidence from this record is therefore that meteorites are not related to the best known star showers. It is fair to presume that the record by months will be somewhat influenced by the times that observers are most abroad. Most of the observations of meteorite falls are made in the northern hemisphere and in this hemisphere observers are more likely to be out of doors and hence more likely to observe the fall of meteorites in the TIMES OF FALL 39 summer than in the winter months. The record shows that as a whole the number of falls recorded is less for the winter than the summer months, yet the number of falls cannot be influenced by that alone since the high record for May and June drops to nearly half that number in July. Further the months of August, September, and October are equally favorable as regards weather for observations of .meteorite falls with those of April, May, and June, yet the latter period 40 30 I 5 I I i I i I I FIG. 9. Curve of meteorite falls by months. much excels in number of falls. The excess of falls in May and June must, therefore, be due to other causes than favor- able conditions of observation and seems to indicate that in the portion of the earth's orbit passed through in these months there is an unusual number of meteorites. The full table up to 1910 for the different months is as follows: Jan. Feb. Mar. April May June July Aug. Sept. Oct. Nov. Dec. 25 24 22 32 44 45 23 36 30 24 24 21=350 This record is shown graphically in the accompanying diagram (Fig. 9). Comparison of the falls of meteorites by months as here given, with those of falling stars and fireballs as given by 40 METEORITES W. H. Pickering,* shows a marked difference of distribution. According to Pickering's list the falling stars and fireballs are much more uniformly distributed through the year than are meteorites, and their period of greatest number is from July to November. In May and June their number is at its minimum. Hence the record seems to show a difference in character between meteors and meteorites and furnishes per se a ground for questioning the gradation that has been supposed to exist between meteors and meteorites. Tabulation of meteorite falls by days of the year seems to show little of significance. The largest number of falls for any one day is 5 on October 13, and this is a month when the total number of falls is not large. Four days show 4 falls each, and 158, or nearly half the total number, no falls at all. The days without falls seem to be scattered indis- criminately through the year, without marked grouping or arrangement. The days showing falls aside from those men- tioned, have from one to three falls each, but do not show any marked grouping. Such a record seems also to indicate that to refer a meteorite falling on the day of a star shower to such showers is unsafe practice, especially if the observa- tions are not sufficient to assign the two to the same radiant. The meteorite falls are so uniformly distributed throughout the year that the two occurrences might easily be coincident without being otherwise related. The full record of the falls by days up to 1910 is as follows: Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. 2. . 3-- 4-- I'-' s'.. 9-- 10. . ii. . 12. . I3-- :i:: Popular Astronomy, 1909, 17, 277. TIMES OF FALL 41 Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. 17 I I 3 i 2 i 18 3 i 2 I I i IQ 2 121 -2 I I . . 2 i *7 20 I I I J 3 2 i I 21 I I 2 I 2 T 22 I 3 i i i 3 i 2 2^ 2 j j i . . i 24 J I I 2 2 i i . . 25 2 i 3 I I I 2 i I 26 .. .. 3 2 I ill 27 I i i 2 I i 3 2 28 I I 2 i 3 29 I I . . I i 3 i 30 I i i i 4 21 . . 2 I i i 23 23 21 29 41 42 21 32 28 24 23 203=27 Of all times of fall of meteorites the most satisfactory for study are probably the hours of fall, since the ratio of num- ber of falls to number of hours is larger than that to days, months, or years. While the hour of fall is not known of as many meteorites as is the year and month, yet of 273 sufficiently satisfactory records are available. Of these 273 falls, 184 occurred in the time from noon to midnight, and 89 from midnight to noon. The record in full is as follows, the total number being less by seven than that recorded for forenoon and afternoon, since of these seven the hour is not known : Hours 12 i 2 3 45 6 7 8 91011 Total A. M i 2 3 2 6 7 7 18 12 10 9 12 = 89 p. M 24 13 19 33 21 15 ii 8 1 6 7 9 3 = 176 This record is shown graphically in the accompanying diagram (Fig. 10). As in the case of the months and the years, it is quite likely that here also considerable allowance should be made for conditions of observation. It is reasonable to expect that the number of falls recorded in the early morning hours would be less than that for other times, since mankind is generally asleep then. That some such allowance must be made is indicated by the records, for the number of falls from midnight to 6 A. M. is only 21, while from 6 A. M. to noon it is 68. From noon to 6 P. M. it is 124 and from 6 p. M. to midnight 60. 42 METEORITES 35 34 33 32 31 30 29 28 27 26 25 24- 23 22 21 20 19 IS 17 16 15 U 13 12 II 10 9 8 7 6 5 4 3 z \ 12 I 2 3 4- 5 6 7 8 9 10 II 12 I 2 3 4 5 6 7 8 9 10 If 12 A M. P. M. FIG. 10. Curve of meteorite falls by hours. TIMES OF FALL 43 The hours of fall are chiefly significant, however, in indicating the direction of movement of meteorites. It will be seen from the accompanying diagram (Fig. n) that all meteorites reaching the earth between noon and mid- night must be moving in the same direction as the earth in its orbit. These are said to have direct motion. Those svm FIG. II. Diagram showing relation of time of day and direction of earth motion to velocities of meteorites. reaching the earth between midnight and noon however, must be moving in a direction opposite to that of the earth or so slowly that they are overtaken by it. Those moving opposite to the earth are said to have retrograde motion. It will be seen that meteorites with direct motion must reach the earth by overtaking it (or being overtaken by it) while those with retrograde motion meet the earth. These differences in direction of motion must produce great differences in the velocity with which meteorites strike the earth, since those overtaking the earth have the earth's velocity subtracted, while those meeting the earth have 44 METEORITES the earth's velocity added to theirs. The earth's velocity about the sun is 18.5 miles per second. All meteorites which move in orbits which are parabolic about the sun have a velocity of 26.16 miles per second. If, therefore, a meteorite having this velocity overtakes the earth it will strike with a velocity of only 7.7 miles per second, its velo- city minus that of the earth. On the other hand a meteor- ite moving in the opposite direction with parabolic velocity and meeting the earth will strike with its own velocity plus that of the earth, or 44.7 miles per second. To these velo- cities must be added that produced by the earth's attraction. It has been shown by Lowell* that this may be as great as 2.66 miles per second and can not be less than 0.53 miles per second. The greatest velocity then at which a meteorite can strike the earth is 44.7+2.7 = 47.4 miles per second, and the least, if the meteorite has direct motion, is 7.7+0.5 = 8.2 miles per second. Such differences in velocities must have a marked effect on meteorites. Meteorites passing into the earth's atmosphere at the higher velocity must be subjected to far greater heat and friction than those moving at the lower velocity. The greater heat and friction would prob- ably be sufficient to burn up all but the largest meteorites, and this, as Newtonf and W. H. PickeringJ have both re- marked, may be the principal reason why so small a number of meteorites is known to fall between midnight and noon. According to the records above given more than twice as many meteorites fall from noon to midnight as from mid- night to noon. This would indicate that most meteorites are moving in their orbits in the same direction as the earth, but taking into consideration the lack of favorable oppor- tunity for observation of meteorites with retrograde motion on account of the time of their fall and taking into consid- eration the greater liability that they will be burned up on account of their greater velocity, it is possible that the dif- ference in quantity of meteorites of the two classes is not as great as appears at first sight. *Science, N. S., 1909, 30, 339. fAm. Jour. Sci., 1888, 3, 36, 10. ^Popular Astronomy, 1910, 18, 264. TIMES OF FALL 45 A study of the table shows that the falls are much more numerous at some hours than others. They are most numerous at 3 P. M., but are also abundant about noon and about 7 A. M. Haidinger in 1867* gave the hours of 178 meteorite falls which may serve for comparison with the above table. Omitting from Haidinger's table about 40 falls which are now known to be unreliable, his results are as follows: 12 i 23 45 6789 10 ii Total A. M i i i 2 3 3 4 10 5 5 5 17 = 57 p. M 7 9 9 16 15 7 5 7 3 o o 3 = 81 Here likewise the afternoon falls are seen to be more numerous than the morning falls, and the number is greater at 7 A. M., n A. M., and 3 p. M. Thus the numbers of falls at different hours seem to retain about the same propor- tion when different yearly periods are compared. On the whole the study of the times of fall of meteorites seems to show (i) that they differ considerably from mete- ors in times of fall, (2) that they are not noticeably related to any of the well known .star showers, and (3) that the rate of their supply to the earth is remarkably uniform. *Sitzb. Akad. der Wiss, Wien, Bd. 55. CHAPTER V METEORITE SHOWERS A striking feature of some meteorite falls (striking in more ways than one), is the fact that a large number of in- dividuals, sometimes thousands, fall at one time and place. Such occurrences are called meteoritic showers, and present phenomena of much interest. These showers have taken place on various parts of the globe and at various times without any seeming regularity or relation. Three of the largest showers, those of Estherville, Forest, and Homestead, took place within the boundaries of the State of Iowa, and three others, Knyahinya, Mocs, and Pultusk, fell in Hungary or the neighboring Poland. The phenomena of violent sounds and brilliant light which usually accompany, the fall of a meteorite are generally intensified in the case of -these showers, though not always to a marked degree. The phenomena attending the shower of Homestead, described on pages 16 to 18 may be con- sidered typical of the more violent form. The distribution of the stones of these showers is usually over an elliptical area with the longest axis of the ellipse in the direction of movement of the meteor (Fig. 12). The greatest distance along which the individuals of a shower have been observed to be distributed in this way is sixteen miles. This was the distribution of the Khairpur shower. The distribution of this and other showers is as follows.* miles Limerick. . . . . 3 miles long Butsura. . . . . 3 miles x 0.6 mile Holbrook... - 3 x 0.6 Pultusk . 5 X I Barbotan . . . . 6 long Homestead. . . 7 x 4 m es L'Aigle - 1% X 2^ Stannern. . . . . 8 x 3 Estherville. . . 8 KlX Pillistfer. . . . . 8 X2*4 Mocs 9 X 2 .Knyahinya. ... 9 m Weston 10 Hessle 10 New Con cord.. 10 Castalia 10 Macao 14 les x 3 long * 3 x 3 lone Cold Bokkeveld. .16 Khairour. . . . 16 RWMg X I X 1 mile miles *Chiefly from Fletcher, Min. Mag., 1889, 8, 225. 46 METEORITE SHOWERS 47 \Vv\\o. FIG. 12. Distribution of the individuals of the Homestead, Iowa, meteorite shower. The shower moved from south to north, the larger individuals being carried farther by their greater momentum. The squares in the diagram represent a square mile each. 48 METEORITES Another feature to be noted in the distribution of the individuals of such showers is that of assortment according to size. The smaller individuals fall at the end of the ellipse nearest the point from which the movement comes, the larger ones at the end farthest away. This difference is probably due to the fact that the greater momentum of the larger masses carries them farther. This rule would seem to be of universal application and any apparent re- versal of it, such as has sometimes been reported, may per- haps be explained as a failure to determine the true direction of movement of the meteor. With a few unimportant exceptions, the individuals of a shower are of the same nature. Single individuals of the Homestead, Stannern, and Pultusk showers were of a some- what different character from the rest but not markedly so. In the Estherville shower gradations from iron-stones to irons were seen. At Brenham also both iron-stones and irons fell. All observed showers have been of stones, but the finding of numerous individuals of iron in single localities such as Toluca and. Canyon Diablo indicates that showers of mete- oric irons sometimes take place also. Finding of stones or irons in large quantity at any lo- cality may be assumed to show the former occurrence of a shower. Showers of stones that have either been observed or found took place at Barbotan, Cronstadt, Estherville, Forest, Futtehpore, Hessle, Holbrook, Homestead, Jonzac, Killeter, Knyahinya, L'Aigle, Macao, Mezo-Madarasz, Mocs, Orgueil, Pultusk, Siena, Stannern, and Weston. Showers of irons occurred at Brenham, Canyon Diablo, Coahuila, Great Nama Land, Imilac, Inca, and Toluca. Numerous other falls, while not producing a sufficient number of individuals to constitute a shower, yet afforded many stones. Thus many stones fell at Admire, Agen, Aleppo, Borgo San Donino, Cold Bokkeveld, Dhurmsala, Kesen, Khairpur, Madrid, Manbhoom, Modoc, Monte Milone, Ness County, Nulles, Ochansk, Sokobanja, Tomat- lan and Toulouse. At Chail, Grazac, Khetree, Jelica, New Concord, Ploschkowitz, and Segowlee from 20 to 40 METEORITE SHOWERS 49 stones fell; at Zsadany 16, at Stalldalen n, at Blansko 8, at Bandong and Lance 6, at Barratta, Bremervorde, Butsufa, and Drake Creek 5, at Harrison County, Marion and Lissa 5, and at numerous other localities 2 to 3 stones. Of irons, about 15 individuals are known from Glorieta; 4 to 6 from Smith ville, Staunton, Steinbach, Trenton, and Youndegin: 3 from Arispe, Bischtiibe, and Crab Orchard; and 2 from Braunau, Chupaderos, Cosby Creek, Hraschina, Losttown, and Tucson. It is highly probable that at many of the above localities not all the individuals which fell were found, so that the numbers would be increased if the full complement were known. The number of stones falling in some of these showers is remarkable. In each of the showers of Pultusk and Mocs more than 100,000 stones fell. In the shower of Holbrook 14,000 stones fell, and in that of L'Aigle 2-3,000. The total quantity of meteoric matter falling in a single shower is also often large though not larger than some single stones. In the Knyahinya shower the stones of which were rela- tively large, over 423 kilos (840 pounds) fell. From Hol- brook 218 kilos were obtained and about the same quan- tity from Pultusk. The question of the amount of area over which meteorites of a shower may be distributed becomes of considerable im- portance when considered in relation to meteorites found. If showers can distribute meteorites over areas covering scores or hundreds of miles, meteorites of similar characters found within such areas should be referred to one fall in- stead of many. This is an especially important considera- tion in regard to the iron meteorites of the class of medium octahedrites, since many of them are separated in point of fall by less than a hundred miles and yet are regarded of distinct origin. Earlier writers were inclined to group into one fall all similar meteorites, even though separated by thousands of miles of distance, but later observations have failed to confirm this view. Until an observed shower can be seen to disperse meteorites for great distances we seem compelled to allow but slight dispersion by a shower. Two 50 METEORITES important meteoric finds, however, seem to be exceptions to this rule. These are Coahuila and Great Nama Land. In the state of Coahuila, Mexico, numbers of meteoric irons of the rare class of hexahedrites are found one or two hundred miles apart. It hardly seems likely that separate falls of these rare meteorites would occur within such a lim- ited area, and the only alternative seem to be to ascribe them to a shower or to assign their distribution to human agency. Fletcher after an exhaustive study concluded that it was highly probable that the usefulness of these masses of iron for anvils and other artificial purposes caused their wide distribution by man. The irons of Great Nama Land are also of a peculiar class, being fine octahedrites. They have been found in various places over an area whose far- thest limits are about fifty miles apart. Here again it seems highly probable that distribution by man has taken place. Opinions differ as to whether the individuals of a shower are separated before striking the earth's atmosphere or come from a single mass broken up in its passage through the atmosphere. Some breaking up of individuals is known to take place in the atmosphere, because individuals show various stages of crust formation on different surfaces. This crust varies from a deep alteration to a mere smoking, and such differences could only arise from successive frac- tures and successively shorter periods of exposure. But the majority of individuals of a shower are thoroughly encrusted on arrival at the earth and are often oriented (Fig. 13). Hence they must have had a nearly uniform period of flight through the atmosphere. As Cohen suggests,* this could only be the case if the breaking up took place simultaneously and very soon after the entrance of the mass into the atmosphere. Now, if conditions favor such a breaking up at the beginning of the atmospheric course, it is not easy to see why meteoric showers are not more abundant, since meteoric stones do not differ greatly in structure and composition. A breaking up of the soft, carbonaceous meteorites would be especially probable. Moreover, while the breaking up of stones in the atmosphere can be con- * Meteoritenkunde, Heft II, 186. METEORITE SHOWERS 51 FIG. 13. Various individuals of the Orgueil, France, shower. Similar numbers indicate the same stone in different positions. Somewhat reduced. After Daubree. 52 METEORITES ceived, it is hard to understand how masses so tough and coherent as the meteoric irons could readily be divided except along a few pre-existing clefts into the great numbers of individuals seen in the Toluca and Canyon Diablo meteorites for example. The principal objections to the view that the individuals of a meteoritic shower are largely separated before reaching the earth's atmosphere seem to be those urged by Daubree, that if the meteorites exist as a swarm in space they should be seen moving as a swarm of lights in the atmosphere, and further that their distribution in falling should be much more irregular and extensive than is found to be the case. So far as the first objection is concerned, it is of interest to note that the Rochester meteorite was described as looking like a "flock of red-hot birds" moving through the air. Nu- merous lights have been seen in the case of other showers. But that the individuals of a shower should be distributed over so narrow an area is remarkable, and shows to what a high degree a fixed direction of movement may have been imparted to the components of a swarm. To sum up, the following facts would lead us to assume that meteorites come within the range of the earth's attrac- tion as a single body and that their disintegration, if any, takes place in the earth's atmosphere: 1. The angularity of most of the individuals of stone showers. 2. The uniform composition of the individuals of a shower. 3. The appearance of meteors in the air as a ball rather than as a swarm of bodies. 4. The narrow distribution of the components of a shower. On the other hand the following facts seem to favor the assumption that meteorites which fall as showers existed as a swarm in space: 1. The complete encrusting of most individuals of a shower. 2. The small number of showers. 3. The regular form of the area over which a shower dis- METEORITE SHOWERS 53 tributes itself and the regular distribution of the individuals over it. 4. The difficulty of breaking up iron masses by atmos- pheric shock. CHAPTER VI SIZE OF METEORITES The largest individual meteorite known is one of the Cape York, Greenland group (Fig. 14). It is an iron meteorite weighing 36^ tons. Its principal dimensions are: Length, 10 feet, II inches; height, 6 feet, 9 inches; width, 5 feet, 2 inches. The meteorite had long been known as a mass of iron to the natives of the region where it occurred, but it had not been seen by white men until Lieut. Peary visited it in 1895. It lay on the shores of Melville Bay, 35 miles east of Cape York, Greenland. The Esquimaux had christened it "Ahnighito," meaning the "Tent," in allusion to its shape and size. About four miles away, lay two other large iron meteorites which were undoubtedly individuals I FIG. 14. Cape York, Greenland, the largest known meteorite. Weight, 36^2 tons. 54 SIZE OF METEORITES 55 of the same fall. All of these meteorites were brought to New York City by Lieut. Peary in 1895 and 1897. The next largest meteorite to "Ahnighito" is that of Ba- cubirito, Mexico (Frontispiece). This is also an iron mete- orite. While it has not been weighed, its estimated weight is 27 tons. Its dimensions as given by H. A. Ward, are: Length 13 feet, I inch; width, 6 feet, 2 inches; thickness, 5 feet, 4 inches. As its shape is much less compact than that of the large Cape York individual, these dimensions are not of much service in comparing the masses of the two bodies. The existence of this meteorite had perhaps long been known to white men, but it was first brought to scientific notice by Prof. Barcena in 1876. Later the meteorite was visited and described by Prof. H. A. Ward. The mass has never been moved from the locality in the state of Sinaloa, Mexico, where it originally fell. Two masses of meteoric iron from Chupaderos, Mexico, which together weigh about 26 tons, must be placed next in the scale of size. Although these two masses were found a few hundred feet apart, the character of their surface showed that they were a single mass before falling. The dimensions of this mass were: Length, 12 feet; width, 7 feet. As separated, one of the masses had about twice the weight of the other. These irons were first located by Europeans as early as 1582 and were removed by the Mexican Gov- ernment to the City of Mexico about 1880. Following these masses in size comes that of Willamette, Oregon, an iron whose present weight is about 15^ tons but the original weight of which was undoubtedly much larger. The dimensions of the mass are: Length, 10 feet, 3^2 inches; breadth, 6 feet, 6 inches; height, 4 feet, 3 inches. The mass is conical in shape and lay for an unknown length of time in a dense forest with its base uppermost. The climate being very moist, conditions were favorable for a rapid oxidation and decomposition of the iron and as a result great cavities (Fig. 15), were formed in the mass which have considerably decreased its original weight. The size of one of these cavities is described by Ward as 3 feet long by 10 to 15 inches across and with an average depth of 16 56 METEORITES inches. Many other such cavities of nearly equal size occur. This mass was moved to New York City in 1906. An iron meteorite of a similar form to Willamette but smaller and little if any affected by decomposition is that of El Morito (San Gregorio), Mexico, the weight of which is about II tons (Fig. 16). The dimensions of this meteor- ite are: Length, 6 feet, 6 inches; height 5 feet, 6 inches; FIG. 15. The Willamette meteorite showing cavities produced by terrestrial erosion and solution. breadth, 4 feet. The existence of this iron was known as early as 1600 and in 1821 a Spanish inscription was cut on it which (translated) read: "Since no one in the world could make it, only God with his power this iron can destroy." About 1880 the meteorite was removed with several others to the City of Mexico. Between this and the meteorite next in size a considerable gap in weight intervenes. The Bendego, Brazil, meteorite which comes in this place, weighs only about 5 tons. It is of a flattened, forked shape, its extreme dimensions being: Length, 7 feet; width, 4 feet; thickness, 2 feet. The meteor- ite is said to have been discovered in 1784 but was not described till 1816. In 1888 it was moved to Rio Janeiro. SIZE OF METEORITES 57 The next largest meteorite in size comes from Australia, and is known as Cranbourne. Several masses occur of this fall, of which the largest weighs nearly 4 tons. It is of rounded form and is now in the British Museum. Next in weight ranks another Mexican iron found not far from those of Chupaderos. This is known as Adargas or Concepcion. It is of flattened form and has the dimensions: FIG. 16. El Morito, Mexico, meteorite. Weight, n tons Length, 3 feet, 10 inches; breadth, 3 feet, I inch; thickness^ i foot, 2 inches. Its weight is about 3 tons. According to an inscription on the iron it was found in the year 1600. It is now in the City of Mexico. The second largest individual of the Cape York fall ranks next in size. This is of conical form and weighs nearly 3 tons. From its shape it was christened, by the Esquimaux, the "Woman." Another Mexican meteorite comes next in size. This is the meteorite of Casas Grandes, which was found carefully wrapped in coarse linen in some ancient ruins in the state of Chihuahua. It is of lenticular form and has the dimensions: 58 METEORITES Length, 3 feet, 2 inches; width, 2 feet, 5 inches; height, I foot, 6 inches. The weight of this meteorite is nearly 2 tons, and it is now in the United States National Museum. The last of known meteorite individuals whose weight exceeds I ton is from Quinn Canyon, Nevada. This is a beautifully oriented, conical iron having the dimensions: Length, 3 feet, n inches; breadth, 2 feet, II inches; height, I foot, 8 inches. Its weight is a little over \]4 tons. It was found in 1908 and is now in the Field Museum of National History, Chicago. The weights of these masses are shown in kilograms in the following table as well as the cities in which the meteor- ites are now preserved. Weight in Where Name Kilograms Preserved Cape York 33> IJ 3 New York Bacubirito 27,500 Mexico Chupaderos, 2 individuals 20,881 Mexico Willamette 14,110 New York El Morito 10,000 Mexico Bendego ' 5,370 Rio Janeiro Cranbourne 3>73 J London Adargas 3>3 2 5 ..... .Mexico Cape York 2,727 New York Casas Grandes !>545 Washington Quinn Canyon 1 A%5 Chicago The above are all irons, and except in one case single masses. Other large iron masses known are those of Magura, weighing 1500 kilos, Zacatecas 1000 kilos, Charcas 784 kilos, and Red River 750 kilos. All of these exceed in size the largest stone meteorite, Long Island, which weighs 564 kilos (1200 pounds). Although broken at the time of its fall this undoubtedly fell as a single individual. The largest unbroken stone meteorite individual known is one of the Knyahinya shower, weighing 293 kilograms (600 pounds). The mass of Bjurbole fell as a single stone weigh- ing about 400 kilos (800 pounds) but it was broken by striking the earth. The iron meteorites will be seen from SIZE OF METEORITES 59 the above statements to far outweigh the stone meteorites in the size of single masses, and this would be expected from the greater resistance to fracture and erosion which their substance is able to exert. None of these large iron masses have been seen to fall. The largest single iron mass seen to fall is that of Cabin Creek, weighing about 100 pounds. From large masses all gradations of size occur .down to material of microscopic dimensions. Some meteoric showers produce large numbers of small stones, others only large ones. In the shower of Holbrook it was estimated that over a thousand individuals of the size of grape seeds fell. Individuals smaller than this are not likely to be found, but it is theoretically certain that they are formed. CHAPTER VII FORMS OF METEORITES The forms of meteorites seem to depend chiefly on the amount of shaping which they undergo in their passage through the earth's atmosphere. This may in turn depend partly on their speed of fall, a lower velocity giving a longer time for shaping. The amount of shaping seems to be independent of the size of the masses, since large and small individuals show similar forms. It is also largely independent of the substance of the meteorites, but there are some forms acquired by iron meteorites which are hardly possible to stone meteorites. Meteorites which break up shortly before reaching the earth present irregular forms such as a rock broken by a hammer might show. A longer course through the atmosphere gives an opportunity for shaping the masses, under the operation of which certain characteristic forms are produced. These forms may be enumerated as follows: Cone-shaped or conoid, shield- shaped or peltoid, shell-shaped or ostracoid, bell-shaped or codonoid, pear-shaped or onchnoid, column-shaped or sty- loid, ring-shaped or cricoid, and jaw-shaped or gnathoid; while among angular forms may be observed cuboidal, pyr- amidal, rhombohedral, tetrahedral, etc., forms. Of the above forms the cone-shaped or conoid is the most common and typical. The cone of such forms is usually low in proportion to its breadth, but its proportions may so vary as to approach the bell shape on the one hand or the shield shape on the other. The form is evidently due to the greater exposure of the forward corners of the fall- ing meteorite to the heat and friction of the atmosphere. These corners, as represented in the accompanying diagram (Fig. 17), are worn away more rapidly than interior portions. From the edges to the center the abrading forces thus grad- ually lessen in intensity and a sloping surface is produced. 60 FORMS OF METEORITES 61 This slope is usually somewhat rounded, and the highest point or apex of the cone is not always situated at the geo- metric center of the figure. It is probably, however, gen- erally in line with the center of gravity of the mass. While FIG. 17. Diagram showing production .of conical form in a meteorite by the greater exposure of its corners. it is true that this conical form may be largely the result of atmospheric shaping, it is also true that a meteorite originally possessing such a form would be turned by the resistance of the earth's atmosphere, as has been shown mathematically by Schlichter,* with its apex foremost. The subsequent action of the atmosphere would then tend sim- ply to preserve this form. *Bull. Geol. Soc. Am., 1903, 14, 112-116. FIG. 18. Front (upper figure) and rear (lower figure) sides of Cabin Creek meteorite. The contrast in the relief of the two surfaces is typical of well-oriented meteorites. FORMS OF METEORITES 63 While the rear side of such a meteorite is much less affected than the front, yet here too some shaping seems to take place, since it is usually more or less concave as com- pared with the convex front side. A marked difference in the character of the pittings is usually also noticeable between the front and rear sides. Those of the front side are small, deep, and oval in outline, while those of the rear FIG. 19. Front side of Goalpara meteorite, showing radial arrangement of pittings. After Haidinger. side are broad, shallow, and more or less circular (Fig. 18). The crust of the front side is thin and dark in color; that of the rear side thick, slaggy, and usually of a reddish or brownish hue. The latter feature shows that the rear side has encountered less air, since it indicates less oxidation. Evidence of the passing of currents of air radially from the apex of the cone is to be seen in the arrangement of the pittings on the front side, (Fig. 19). These pits are as a rule elongated, oval, and furrow-like, and broadest on the side toward the edge. The slope of their sides is commonly 64 METEORITES unequal and the position of the steeper slope is not constant. The pits do not as a rule merge into one another on all sides, but follow lineal and radial courses. In size the pits are usually larger on iron than on stone meteorites, their aver- age range being from one-fourth to one inch in diameter. The pits are not usually to be found on the apex of the cone, the surface there being characteristically smooth. A char- acteristic feature of the edge where the front and rear sur- faces of the meteorite join is a thickening of the crust caused FIG. 20. Jonzac meteorite, showing lateral edge produced by meeting of currents from front and rear. by an accumulation of fused matter (Fig. 20). This crust is also often notably blebby in character. The rear side also often exhibits adhering particles which have the ap- pearance of being fragments of crust. Haidinger regarded these as fragments which accompanied the meteorite from space with a velocity equal to it and fused upon it when its speed was lessened. Rath and Tschermak, however, thought them fragments from the front side of the mete- orite which were thrown to the rear and fused upon that surface by hot air streaming into the space behind. Many large meteorites, and especially iron ones, exhibit the conical or conoid form in a high degree. Among such large FORMS OF METEORITES 65 iron meteorites may be mentioned El Morito. Willamette, (Fig. 21), Quinn Canyon, and Cabin Creek. The iron meteorite of Cabin Creek weighs about 100 pounds. Its front or apical side is covered with numerous, deep, elongated impressions one-half to one inch in diam- ^Bfc*, FIG. 21. A conoid or cone-shaped meteorite. Willamette. Weight, about 15 tons. eter. The apex is free of crust, but from it fused threads of the substance of the meteorite run radially. These threads are hair-like in thickness and from one to four inches long, and may be traced, according to Kunz, on the slope and bottom of the pits. The rear face is relatively flat and shows broader, shallower pits than the front. These pits 66 METEORITES are from one to two inches in diameter. The rear surface has a rough, scale-like crust about one millimeter thick. Brezina regards the thin, smooth crust of the front side of this meteorite as giving evidence that it was in the state of a thin, mobile liquid, while the thick crust of the rear side shows that it was in a viscid condition and, therefore, must have received less heat. Other smaller iron meteorites exhibiting a more or less FIG. 22. A conoid or cone-shaped meteorite. Long Island, Kansas. Stone, weight about 1 100 pounds. The symmetrical arrangement of the pittings is noteworthy. well-marked conoid form are Braunau, Carlton, Cleveland, and Costilla. Of stone meteorites the largest known to exhibit the conoid form is Long Island (Fig. 22), the weight of which when entire was about 1400 pounds. Here a smooth apex, deep, radial pitting of the front side, and broad, shallow pitting of the rear side are exhibited. The altitude of the cone of this meteorite is 20 inches, and the greatest diameter of the base 34 inches. Another. large stone meteorite exhibiting the same form is the large mass of Bath Furnace (Fig. 23). This is an FIG. 23. Side (upper figure) and front (lower figure) views of the Bath Furnace meteorite. This is a well-oriented stone meteorite weighing about 180 pounds. The symmetrical arrangement of the oittings is well shown. 68 METEORITES individual weighing 180 pounds and covered uniformly with a black crust. On the front side appears the usual smooth apex with rows of pittings radiating from it. The rear side is relatively smooth. Variation of the conical shape produced by a diminution of the altitude of the cone gives, as has been stated, shield- shaped forms. Of this form the N'Goureyma and Algoma meteorites furnish excellent examples. Both of these are iron meteorites. The N'Goureyma meteorite is 22 inches long and n inches broad, and its greatest thickness is 3^ inches. Its outline seen broadside is very irregular and the boss of the shield is placed near one end. The front or boss side is convex and marked by small, deep, rounded pits, the walls of which often exhibit smaller pits giving them a pockmarked appearance. Fine, furrow-like depressions also give this surface a scale-like semblance. Contrary to the usual rule, the crust is rougher and darker on the front than on the rear side. The pits of the rear side are large, shallow, smooth, and elongated. Cohen was of the opinion* that the original form was more symmetrical, but that it was strongly modified by the erosion of the air. On account of drift markings seen on both surfaces he also concluded that the meteorite moved through the air at an acute angle to the direction of its motion. Hobbs,| however, urged that the drift markings on the rear side were plainly the result of air currents forced through two holes in the meteorite, since they were found only in the area peripheral to these openings. It is highly probable, therefore, as Hobbs concluded, that the meteorite took the broadside attitude in its flight. The outline of the Algoma meteorite (Fig. 24), on its broadside is roughly elliptical with axes of 10 and 6 inches. Its thickness varies from about one inch near the geometric center, to knife edges at several points. A few large, shallow pits occur upon the front side, but a more remarkable fea- ture is a complete series of radial furrows extending over the surface from the center outward. These are knife-like edges from one-fifth to one-tenth of a millimeter in width at the *Am. Jour. Sci., 1903, 4, 15, 254. fBull. Geol. Soc. Am., 1903, 14, 108. FORMS OF METEORITES 69 base, separated by furrows from one to two millimeters wide. The ridges are modified somewhat in their course by the structure of the meteorite, but in general pursue a rectilinear direction with a slight curve to the left. On the rear side the surface is concave and a number of broad, shallow pits ap- FIG. 24. Front and side views of Algoma, a peltoid or shield-shaped meteorite. An iron meteorite. Weight, 9 pounds. pear but the ridge-like markings of the front side are entirely absent. Somewhat similar in form to the shield-shaped or peltoid meteorites are the shell-shaped or ostracoid meteorites but the origin of the ostracoid form is probably quite different from that of the peltoid form. Ostracoid meteorites are thin and extended, concave on one side and convex on the other, with much more marked curving than in the peltoid meteorites. Instead of owing their form chiefly to atmos- 70 METEORITES pheric shaping, the ostracoid meteorites are probably pri- marily shaped as a scaling from some larger body. Many of the Canyon Diablo meteorites showthis shape to a marked degree. The best illustration among stone meteorites is Butsura, which fell in several pieces which, when put to- gether, made a well-marked shell shape. Such a shape is ill adapted to withstand atmospheric resistance, and hence rup- ture of the individual is likely to occur. If the apex of the cone be raised, and the side slope be concave, a bell-shape or codonoid form will be produced, of which several meteorites afford illustrations. One of the best shaped of these is the Durala meteorite. The height of this meteorite is 7 inches; the diameter of its base 10 inches. The outline of the base is triangular rather than circular, but the angles of the triangle are considerably rounded. The surface of this meteorite is almost uniformly smooth and shows little or no contrast in appearance be- tween front and rear sides. By being elongated still more in the direction at right angles to its circular outline the bell-shaped or codonoid form passes into the pear-shaped or onchnoid form. Only iron meteorites, so far as known, exhibit this shape. Among these Boogaldi and Charlotte furnish excellent examples, (Fig. 25). In meteorites of this shape the orientation changes. The large, heavy, blunt end is now foremost, the small, pointed end at the rear. That such is the position of these meteorites in falling is shown beyond a doubt by the markings on Boogaldi. At the thick, heavy end of this meteorite, well-defined concentric zones of fused oxides may be seen, with transverse furrows running in the direction of the thinner end of the meteorite. The disappearance of both zones and furrows is gradual and in the same direc- tion (Fig. 26). Liversidge* regards these zones of oxides as thrown up by the resistance of the air "just as waves are formed in water or sand by the wind or at the bows of a boat." At the small end of the meteorite longitudinal ridges and furrows may also be seen in a "skin" of fused oxides. These have the same direction as the furrows at *Proc. Roy. Soc. N. S. Wales, 1902, 26, 343. FORMS OF METEORITES 71 the larger end, an'd there are remains of drops where the melted material dripped off at the small end. Another indication that the meteorite moved large end foremost, although this evidence is not always conclusive, lies in the fact that it was resting on this end when found. The length of the meteorite is 5 inches and its diameter at the large end FIG. 25. Onchnoid or pear-shaped meteorites. Charlotte at the left, Boogaldi (from a cast) at the right. Both are iron meteorites. Charlotte weighed 9 pounds; Boogaldi, 5 pounds. 3 inches. The surface in general is smooth and shows no pittings except for the furrows referred to. The continuity of the etching figures to the edges of the meteorite as seen in section (Fig. 27) shows that the form of the meteorite is due to erosion. The Charlotte meteorite is about the size and shape of Boogaldi, but somewhat more flattened laterally, and one side is concave. No markings such as those which so dis- 72 METEORITES FIG 26. Forward end of the Boogaldi meteorite showing waves and ridges formed on its surface by fusion during its fall. Enlarged 1.7 diameters. After Liversidge. \ Fig. 27. Etched section of the Boogaldi meteorite. The continuity of the etching figures to the edges of the mass shows that the form of the mete- orite has been produced by erosion. Enlarged 1.3 diameters. After Liversidge. 74 . METEORITES tinctly orient the Boogaldi meteorite seem to have been observed upon Charlotte, but it is highly probable that its position in falling was likewise with the large end foremost. If the meteorite is still more elongated and acquires a somewhat convex instead of a concave curve in the direc- tion of its length, a column-shaped or styloid form like that of the Babb's Mill meteorite (Fig. 28) will be produced. The length of this meteorite is 3 feet; breadth 10 inches, and thickness 6 inches. It weighed 290 pounds. It was thought by Blake, who originally described it, that this FIG. 28. Babb's Mill. A styloid or column-shaped meteorite. Length, 3 feet. Weight, 290 Ibs. meteorite was a residual nodule of an irregularly shaped mass from which the irregular portions had been thrown off by terrestrial weathering, but it seems quite as likely that the form was acquired in falling. Of ring-shaped or cricoid forms among meteorites but a single example seems to be known, that of Tucson (Fig. 29). This meteorite is in the form of a metallic ring, the exterior diameter of which varies from 49 to 38 inches and the interior diameter from 26^ to 23 inches. The width of the thickest part of the ring is 17^ inches and of its narrowest part 2^ inches. The greatest thickness at right angles to the plane of the ring is 10 inches. It will thus be seen that the ring is somewhat irregular in form, but a general ring shape is well exhibited. There are no oriented pittings upon the ring. As to the origin of this ring, opinions differ. Haidinger concluded* that the meteorite rotated in its *Sitzb. Wien. Akad., 1870, Bd. 61, II, p. 506-511. FORMS OF METEORITES 75 descent and that thus a hole was bored through it by the air. Observation of the action of the air upon other meteo- rites does not confirm this view, however. It is more prob- able that the ring existed preterrestrially as a portion of an otherwise stony mass and that the stony portion fused or fell away in the passage of the mass to the earth. This view is rendered somewhat more probable by the fact that the iron contains about five per cent of silicates. FIG. 29. A cricoid or ring-shaped meteorite. One of the Tucson, Arizona, fall. It is an iron meteorite and weighs 1514 IBs. Of jaw-sKaped or gnathoid meteorites Kokstad and Hex ,River (Fig. 30) furnish excellent examples. These are both iron meteorites and of nearly the same size. Kokstad is 26 inches long, 12 inches wide at the angle of the "jaw," and 3 inches thick. Its surface is comparatively smooth except for one large circular depression probably caused by the fusing out of a troilite nodule. Hex River is 20 inches long, II inches wide, and 7 inches thick. It is thus somewhat 76 METEORITES more massive than Kokstad and approximates the pear shape. Unlike Kokstad, it is deeply pitted all over its surface. The pits are broad, shallow depressions of rather uniform size, in part so arranged as to give the impression of furrows passing around the "jaw" at right angles to its length. Nothing in the distribution, size, or shape of the pittings seems to give a clue as to the position of the mete- orite in falling. From its resemblance in form to the pear- shaped meteorites, however, it may be surmised that it FIG. 30. Gnathoid or jaw-shaped meteorites. Upper one, Hex River; lower, Kokstad. Both are iron meteorites and weiirh about 100 pounds each. fell with the heavy end foremost. The jaw shape is ex- hibited in some degree also by the great Bacubirito mete- orite, (Frontispiece) although the form is not as well marked as in those previously described. The surface of this mete- orite is quite uniformly covered with pittings, regular in size, 2 to 3 inches in diameter, with well-defined walls and quite shallow. No characters seem to afford data for an orienta- tion of the mass. A single, small specimen of Toluca in the Field Museum collection also exhibits a gnathoid shape. It is reasonable to suppose that this shape would be exhib- FORMS OF METEORITES 77 ited only by iron meteorites, since stone meteorites if of this form would be likely to be broken up in passing through the earth's atmosphere. The most satisfactory suggestion that seems to have been made regarding the origin of the gnathoid shape is that of Brezina,* who thought that it arose from the breaking apart of a ring like that of Tucson. It is clear that such forms might arise from a dismembered ring, but whether this was the actual origin of those known cannot, of course, be stated positively. The other meteorite forms mentioned, such as cuboidal, pyramidal, tetrahedral, etc., are exhibited by various meteorites and especially those falling in showers. Here the action of the air has frequently not been sufficient to greatly modify the original angular form, and hence such forms as would be found in freshly broken terrestrial rocks occur. *Verh. der K. K. geolog. Reichsanstalt, Wien, 1887, 289. CHAPTER VIII CRUST OF METEORITES All meteorites are characterized by a more or less smoothed or rounded coating differing in color or texture or both from that of the substance of the meteorite beneath. This superficial skin or coating is known as the "crust" and is a distinguishing character of meteorites. It is obviously the result of fusion of the surface of the meteorite by heat encountered during the passage of the mass through the atmosphere. Being the result of fusion, it varies according to the constitution of the meteorite. If the meteorite is composed of feldspar and augite, for example, which are minerals fusible with comparative ease, a smooth and varnish-like crust flowing in little rivulets over the surface is seen. If, however, the meteorite is composed chiefly of the difficultly fusible minerals, bronzite and chrysolite, as are the majority of stone meteorites, a rough, scoriaceous crust is formed. The color of the crust also varies with the composition of the meteorite. Meteorites containing iron compounds in even small quantity have a black or dark-colored crust on account of the presence of these sub- stances. If, however, iron compounds are lacking, the crust may be nearly colorless as in the meteorite of Bishopville, or yellowish, as in Bustee. Inasmuch as iron or its com- pounds is a nearly constant component of meteorites, a black or dark crust is to be found on the majority of them. Since iron meteorites consist almost wholly of iron, the crust upon them is usually black when they are freshly fallen. Iron meteorites are, however, not as uniformly or plainly encrusted as are the stone meteorites. The crust of iron meteorites when it can be separated is found to have the composition of magnetite. It is usually very thin (i mm. and less) and dull or only weakly shining. Its surface may be smooth or rough, shagreened or warty, and occasionally 78 CRUST OF METEORITES 79 slaggy. Reichenbach called it "iron glass" (Eisenglas), but the term is hardly an apt one, since, as Cohen remarks, the substance is not iron and not glass. In addition to the formation of crust, there is an alteration by heat of the periphery of iron meteorites producing a zone exhibiting a granular structure which may be seen about the edge of the meteorite. It is illustrated in the accompanying figure I FIG. 31. Heat crust of Charlotte meteorite. The heating of the surface in the passage of the meteorite to the earth produced the granular edge seen at the left. X2.. of a section of the Charlotte meteorite, (Fig. 31). This zone varies in width from an extreme of 10 mm. in Cabin Creek to less than I mm. in Prambanan. As a rule, the zone is thinner the greater the weight of the meteorite, although there are some exceptions to this. Such a rela- tion of the width of the zone to the weight of the meteorite is doubtless due to the greater difficulty of heating the larger masses. Since most iron meteorites have not been 80 METEORITES found until some time after their fall, they have generally suffered oxidation which has considerably altered if it has not altogether destroyed their fusion crust. In its place one finds, as a rule, an oxidized crust, usually of a reddish brown color and of varying thickness. The rapidity with which it may form is largely dependent on the constitution of the meteorite, those having considerable chlorine oxidizing FIG. 32. Surface relief of Juncal meteorite produced by terrestrial erosion. the most rapidly. Climate is also an important factor in determining the rate of oxidation of iron meteorites. De- composition takes place much more rapidly in wet than in dry climates. The erosion of dry climates sometimes produces peculiar markings on the surface of iron meteorites. These markings may simply be numerous, small, circular pits independent of one another or they may run together to produce rills like those seen on the surface of the Juncal meteorite (Fig. 32). The position of these rills coincides with the lamellar struc- ture of the meteorite. Such an appearance is especially CRUST OF METEORITES 81 characteristic of the meteorites which have been obtained from the Chilean desert, but it is also shown on the Austra- lian meteorite of Youndegin. In both stone and iron meteorites the thickness and other characters of the crust vary in different parts of the mete- orite. Many of the characters depend on the position of the meteorite in its flight through the atmosphere and the length of time during which the different surfaces have been exposed. On the side of the meteorite which was in front in its course the crust is thinner and shows more complete fusion than on the side which was behind. Other portions of the meteorite may show a thickened crust from the flow of fused matter to those points. A feature occasionally seen in the crust of some stone meteorites is a marking off by fissures into little angular fields such as are seen on "crackled" earthenware. These are evidently due to con- traction in the cooling of the crust. In contrast to the generally rough appearance of the crust of some stone meteorites, there occur in places spots which appear to have been glazed over. These are generally round or oval, from 2 to 10 mm. in diameter, smooth and usually of a yellowish or reddish color as compared witb the black crust. They appear to represent the location of exceptionally fusible con- stituents upon the surface of the meteorite. They are quite characteristic of the stones of Mocs, also are seen in some of those of Modoc, L'Aigle, etc. Where nickel-iron appears upon the surface of stone meteorites it generally projects in rounded forms, especially if the grains are large. This indicates that it is more refractory than the siliceous constituents of the meteorites. In other cases, however, the nickel-iron may oxidize sufficiently to form with the silicates a more fusible compound and cause pits instead of knobs. The thickness of the crust of the stone meteorites depends somewhat upon their texture, a compact stone having a thinner crust than one of more open texture. Upon meteorites of a relatively porous character the crust may reach a thickness of 10 mm., though a lesser thickness is usual. The stones of the meteoric shower of Mocs fur- nish proof that in the stone meteorites the thickness of the 82 METEORITES crust does not vary with the size of the stone. Though the stones were of many sizes, the thickness of the crust was invariably ^3 to ^ mm. All of these features indicate that the crust is the result of a sudden, brief heating. When seen under the microscope in section, the crust of most stone meteorites presents the interesting feature of three, and in some cases four, well-marked zones (Fig. 33). FIG. 33. Microscopic section of a Mocs meteorite showing (above) three zones of the crust. X 70. The outermost of these, called the fusion zone, is thin as compared with the other two, glassy, black and opaque to brown and transparent. Beneath it lies a broader, trans- parent zone in which the constituents of the meteorite ap- pear little, if any, changed. This was called by Tschermak the absorption zone (Saugzone). Next and last follows a broad zone of black, opaque, spotted appearance. This zone may be so broad as to make up four-fifths of the width CRUST OF METEORITES 83 of the crust. The constituents of the meteorite appear in this portion of the crust to be in normal condition, but impregnated with black, generally opaque matter. Hence this zone was called by Tschermak the impregnation zone (Impragnationszone). The relative width of the three zones was found by Ramsay in Bjurbole to be as follows: Outer zone, o.i mm.; middle zone, 0.2 mm.; inner zone, 0.3 mm. The relative and total widths of these zones vary in differ- ent meteorites, being usually the broadest in the more fri- able and porous meteorites. In compact meteorites the crust is thinner, and often only the outer, fused zone can be distinguished. In the crust of other meteorites again the middle, transparent zone may be lacking and only the opaque, outer and inner zones be seen. On many of the stones of Mocs, Brezina noted a very thin layer of yellow transparent substance outside of the fusion zone, so that the number of zones was raised to four. The origin of the three crust zones seems best accounted for by the theory of Tschermak that fused matter from the exterior penetrates through the middle zone and congeals in the inner zone. In the crust of the St. Michel meteorite Borgstrom* found a large excess of pyrrhotite in the inner or impregna- tion zone. As pyrrhotite is the most fusible ingredient of this as well as of all meteorites, he urged that the opacity of the impregnation zone was probably due to the chilling and collecting there of pyrrhotite which had flowed inward from the outer crust in a fused condition. From the fusing temperature of pyrrhotite Prof. Sundell calculated the length of time during which the crust had been exposed to heat, or in other words the time of formation of the crust to be 1.16 seconds. Between a completely developed crust and no crust at all, various gradations occur according to the length of time during which a meteoritic surface is exposed during the flight of the meteorite through the atmosphere. As meteoritic in- dividuals often break up during their flight to the earth, different surfaces will be exposed for varying lengths of *Bull. Com. Geol. de Finlande, 1912, 34, 22. 84 METEORITES time. Those which become exposed just before reaching the earth will have no crust at all; those exposed a little longer will appear as if smoked. A little longer exposure may produce a blackening of the entire surface, but without glazing or smoothing, and between this and a well-formed crust all gradations may occur. Such a partial encrusting is often called a secondary crust to distinguish it from the well-formed primary crust. Individuals of a meteoritic shower frequently exhibit several grades of secondary crust, due to successive disruptions in the air. CHAPTER IX VEINS OF METEORITES Many of the stony meteorites are penetrated by black, thread-like veins. These veins may run continuously or with interruptions, and may be close together or scattered. When very numerous they give to the meteorite a brecciated appearance and may be so abundant as to color the whole mass. They seem to be largely confined to the chondritic meteorites, being seen among the achondrites only in Bish- opville. So abundant are they among the chondrites that Brezina made the presence of veins a ground of subdivision of all their groups. Of 267 chondrites he reported 145 either veined or breccia-like. The white and intermediate chon- drites showed the largest per cent of veining, and the gray chondrites the largest amount of breccia-like structure. In the spherical and crystalline chondrites both features were less frequently observed. The course of the veins is not as a rule marked by any particular system or direction. It is generally more nearly straight than curved, but there may be much forking and anastomosing. In the Bluff meteorite two systems of veins cross at angles of about 45. The narrower of these veins is of nearly uniform width and was observed over a plane 4x15 inches. The other vein varies in width and is less ex- tensive. Some veins are so delicate as to appear in section like the finest hair, scarcely o.oi mm. in thickness. On the average they have a thickness of about o.i mm. The thick- ness as a rule is pretty uniform, but swellings and knottings may appear. Sometimes vein-like masses are seen more than an inch in thickness, and single stones of Pultusk and Mocs seem to be made up wholly of vein material. Other meteorites showing large vein masses are Chantonnay, L'Aigle, Orvinio, and Stalldalen. Another feature allied to veins seen in many meteorites (best disclosed on breaking 85 86 METEORITES the stone) is that of smoothed, blackened surfaces passing in various directions. They are termed armored surfaces (H^rnischflache). Such surfaces have been likened to the slickensides of terrestrial rocks, also to stylolites. They pass to veins by insensible gradations. FIG. 34. Cross section of a vein of one of the Mocs meteorites. The vein mass appears as a broad black band through the center. It is in part intermixed with the adjoining ground mass and in part has well-defined walls. The gray spots are lumps and spheres of nickel-iron illuminated by reflected light. The branching of some of these into clefts, one of which is still open, is of interest. X2O. After Tschermak. Under the microscope the principal substance of the black veins appears like that of the first and third layers of the crust. It is opaque, dull, half glassy, showing in reflected light fused spheres of nickel-iron and pyrrhotite and grains of nickel-iron branching out into delicate leaves (Fig. 34). VEINS OF METEORITES 87 The occurrence of these fine iron veins furnishes the chief distinction between the substance of the crust and that of the black veins. The boundary between the veins and the adjoining stone is at times sharp and again gradual. In the latter case the ground mass seems to be impregnated by a black, half-glassy injection. In the broader veins a distinct flow structure of the substance is evident. The finer veins in their course tend to avoid the chondri and follow the pyrrhotite. In three meteorites analyses of the normal stone and of the vein substance have been made. These meteorites are Orvinio, Stalldalen, and Bluff. The analysis of the nor- mal stone is shown in the following table in each case under a, that of the vein under b. ANALYSES OF VEIN MATERIAL a b a b a b Si0 2 38.01 36.82 35.71 38.32 37.70 38.96 A1 2 O 3 2.22 2.31 2.11 2.15 2.17 1.89 FeO 6.55 9.41 10.29 9.75 23.82 22.98 MgO 24.11 21.69 23.16 25.01 25.94 7-52 CaO 2.33 2.31 1.61 1.84 -2.20 tr. Na 2 O 1.46 0.96' .... K 2 O 0.31 o . 26 ' MnO .... .... 0.25 i.oo .... .... NiO .... 0.20 0.42 ' Fe 22.34 22 -H 21.10 17.47 4-4i 2.30 Ni+Co 2.15 3.04 1.78 1.02 1.75 3.26 1.94 2.04 2.27 2.51 1.30 0.26 P 2 O 5 0.30 0.31 101.42 100.95 98-7 8 99-8o 99.29 97.17 G= 3-675 3-6oo 3.733 3.745 3.510 3.585 REFERENCES 1. Orvinio. Sipocz. Sitzb. Wien Akad, 1874, 70, i, 464. 2. Stalldalen. Lindstrom. Geo!. Foren. Stockholm, 1878, 4, 53-54. 3. Bluff. Whitfield and Merrill. Am. Jour. S.ci., 1888, 3, 36, 119. These analyses make it evident that the substance of the veins does not differ essentially from that of the mete- orite. The vein matter, therefore, has doubtless been formed by alteration in place of the substance of the mete- orite and does not represent foreign matter introduced into a fissure, as is the case with most veins in terrestrial rocks. For this reason, therefore, the term veins applied to these 88 METEORITES formations in meteorites is somewhat misleading, and the lack of analogy to veins in terrestrial rocks should be kept in mind. The similarity in composition between the vein and the substance of meteorites and the resemblance of the vein matter to the crust seem to make clear the origin of the veins. They are apparently produced by the penetration of heat into the fissures of the meteorite during its passage through the atmosphere. Some have thought that the vein matter was fused matter from the surface which flowed into fissures; but, as Tschermak pointed out, the low tem- perature of the interior of the meteorite would probably prevent this. In Chantonnay he found several fissures into which the fused matter from the surface had penetrated to a depth of but 6 mm., leaving the fissure open for the re- mainder of its course. Earlier writers were inclined to re- gard the veins of preterrestrial origin, but there seems no need to assume this. Reichenbach regarded some veins of cosmic, others of telluric, origin. The cosmic veins, he stated, avoided the larger constituents of the meteorites and characterized the iron-rich, dark, and compact chon- drites. The telluric veins passed through the larger constit- uents and characterized the light, iron-poor chondrites. These distinctions have not been accepted by later ob- servers. That the substance of the chondrites turns black upon heating has been abundantly proved experimentally. Meunier observed that a piece of Pultusk became black upon heating it in a stream of carbonic acid gas. A piece of Tadjera was turned red by heating in a current of air and later black by heating in hydrogen or carbonic acid gas. Cohen, heating pieces of Lancon in a platinum boat to a point at which hard glass softened, obtained only a reddish brown color both in a current of air and in one of hydrogen; but on heating in the flame of a blast lamp in a platinum crucible, a black color appeared. He, therefore, concluded that the amount of heat rather than the presence or ab- sence of oxygen caused the production of the black color. Besides black veins, metallic veins consisting largely of nickel-iron occur in some meteorites, especially in Farming- VEINS OF METEORITES 89 ton and Tabory. The form and distribution of these metallic veins seem to be similar to that of the black veins, and it is the belief of Cohen * that all gradations occur between the two kinds. The broad, black veins often have more or less metallic interiors and an increase of this metal- lic substance would produce a metallic vein. The present writer, has however, f called attention to the fact that a diminution of the metallic substance of an iron-stone mete- orite might equally well produce the veins. Perhaps both methods of origin are possible. In the Farmington mete- orite, which is the one in which the metallic veins are best developed, the veins penetrate throughout the mass, and it is not easy to understand how a substance so difficultly fusible as nickel-iron could be so distributed by the pene- tration of heat into the fissures. Black veins also occur in iron meteorites, penetrating between the bands and at times crossing their course. Their usual contour has been accurately described by Cohen as like that of a stroke of lightning. The substance of these veins where it has not suffered terrestrial alteration is hard, takes a good polish, and resists acid. It was regarded by Reichenbach and Brezina as magnetite and as probably pro- duced by the penetration of heat and air into the fissures of the meteorite. In the view of the present writer, a fissuring of iron meteorites can take place after their fall by the penetration of slow oxidation inward, which would give a vein-like appearance. The alteration of a magnetite vein to limonite might also produce such an effect. Structures similar to the armor faces of stone meteorites have been noted in the iron meteorites of Quesa and Sacramento. Probably related to armored surfaces (page 86) are the slickensides seen in some stone meteorites, notably Long Island (Fig. 35). These in Long Island lack the dark color, as if rubbed with graphite, which is common to the average armored surface. They are smooth, shining, somewhat un- even, and striated in the direction of movement. The sur- faces may be parallel to each other at different levels and *Meteoritenkunde, Heft II, p. 121. fAm. Jour. Sci., 1901, 4, n, 60-62. 90 METEORITES also run in different directions. In Long Island three of these directions are nearly at right angles to one another. The movement has brightened and elongated the metallic grains, but produced no other changes in the immediately adjoining areas. Cohen has urged that these surfaces are FIG. 35. Slickensided surface, Long Island meteorite. Natural size. terrestrial in origin, but this seems to the author unlikely. Their coursing in different directions makes it difficult to ascribe them to impact upon the earth, as was done by Cohen, as in such case the movement would be expected to be in a single direction. In iron meteorites evidence of internal movement is afforded by faulting in the etching figures. Bridgewater, Carlton, Descubridora, Magura, and Puquios are meteorites VEINS OF METEORITES 91 in which such faults have been described. In Puquios dislocation has taken place in several directions. The largest fault extends the entire length of the mass and has a throw of 3 mm. In Descubridora a throw of from 6 to 12 mm. has been noted. The faulting is believed by Brezina to be due to the impact of the meteorite on the earth, though Howell was inclined to ascribe it, in the case of the Puquios meteorite at least, to the passage of the mass near the sun, causing high heating. Another evidence of movement in iron meteorites' is afforded by bent or curved figures such as have been noted in Bacubirito, Carlton, Glorieta, Jamestown, and Toluca. Such figures are usually confined to a small area on a single meteorite and are probably correctly assumed to be produced by the impact of the meteorite on the earth. The writer has produced them by boring or hammering a meteorite. CHAPTER X STRUCTURE OF METEORITES STRUCTURE OF IRON METEORITES Iron meteorites as seen upon a polished surface usually present a homogeneous and uniform structure. Some- times cleavage planes of considerable dimensions pass through individual masses, and broken fragments of iron meteorites often exhibit a hackly surface or various small cleavage planes. Large structural peculiarities are, how- ever, generally wanting. The nearest approach to them is to be seen in the iron of Mount Joy, which appears to be composed of irregular nodules, some of which are an inch or two in length. These are generally regarded, however, as phases of a crystalline structure on a large scale. But though large structural features are wanting, when examined intimately most iron meteorites exhibit a well-marked minute structure. It is most successfully brought to view by etching a polished surface of the meteorite. Then there appear, on the majority of iron meteorites, figures formed of parallel bands intersecting in two or more directions, (Fig. 36). These are called Widmanstatten * figures after Alois von Widmanstatten of Vienna, who first produced them in 1808 by heating a section of the Agram meteorite. The presence of these figures shows that the iron is in reality made up of a number of laminae or plates lying in parallel and crystalline positions. Wherever the structure appears in this form it is found that the plates are parallel to the four pairs of faces of an octahedron. The structure is therefore octahedral, and such irons are known as octahedral irons. While the structure may be in a general way de- scribed as made up of plates parallel to the planes of an octahedron, it is in reality as a rule more complex than this. Generally each plate or lamella is itself made up of many *The word is sometimes given the adjective form, Widmanstattian. 92 STRUCTURE OF METEORITES 93 smaller plates combined in twin position, arranged accord- ing to the positions of twelve faces of the trisoctahedron, 112. These smaller plates or lamellae repeat the structure of the larger lamellae. So far as their coarse structure is con- cerned, the larger lamellae as seen in section consist of a FIG. 36. Typical octahedral etching figures of an iron meteorite. The Red River meteorite. broad, central band to which Reichenbach gave the name of kamacite, from /ea^a, a shaft, bounded on either side by a thin border to which Reichenbach gave the name taenite, from rcuvia, a ribbon. Angular interstices between intersecting lamellae may be filled with a homogeneous substance known as plessite or partly with this and partly with structures repeating on a smaller scale those of the larger lamellae. The interstices are usually known as fields. 94 METEORITES Various minute structures may be scattered through the fields, most prominent among which are combs (Kamme). These run out from the principal lamellae, but differ from the primary lamellae in their smaller size and also in that the taenite and kamacite in them are fused together. Other in- clusions in the fields may be (i) minute flakes of taenite FIG. 37 FIG. 38 FIG. 39 FIG. 40 FIGS. 37-40. Four figures showing how the etching fig- ures of an octahedral meteorite will be affected by the direction in which the section is cut. giving a shimmering appearance and (2) slightly indicated small lamellae giving a half-shaded effect. Again small la- mellae may lie nearly alone in a field. Considerable differences in the figures of different mete- orites are produced by variations in the grouping, size, and shape of the lamellae. Thus the lamellae may be in par- allel groups of nearly equal length or they may be of un- equal length, the middle ones in this case usually being the longer. Again the lamellae may be long or short and may STRUCTURE OF METEORITES 95 be of uniform width or have rounded or irregular outlines. Again, they may vary much in width, extremes of variation being from a fraction of a millimeter to several millimeters. Such variations as are above described do not occur to any large extent on any single meteorite. Almost without ex- ception, the figures are uniform throughout any individual mass and for all the individuals of a single fall. This fact aids in distinguishing meteorites of different falls. The an- gles at which the lamellae intersect in any given section will depend as shown in the accompanying figures (Figs. 37-40) on the direction of the section with relation to the octahedral structure. Thus if the section should be par- allel to an octahedral face (as in Fig. 37) the section will show three systems of bands intersecting at angles of 60. If it is parallel to a cubic face (Fig. 38) it will show two systems of bands intersecting at angles of 90. If it is par- allel to a dodecahedral face it will show two systems of bands intersecting at angles of 109 28', and two others which will bisect this angle (Fig. 39). If it should be made in any other direction, and this is of course most likely, it will show bands running in four directions and intersecting at unequal angles (Fig. 40). A modification of the octahedral structure observed by Rinne* in a section of the Bethany (Gibeon) iron, showed, as may be seen in the accompanying figure (Fig. 41), in addition to the usual octahedral lamellae, lamellae running parallel to the planes of a cube. To meteorites having this structure Rinne gave the name tesseral-octahedrites. In English, the term tesselated octahedrites is perhaps better. Plates of pyrrhotite found in this iron followed the planes of the dodecahedron in their arrangement. Another of the Bethany (Mukerop) irons showed a struc- ture which has not been elsewhere observed which seems to be a mass twinning. The meteoritic individual which shows this twinning weighs about 350 pounds. It appears on etch- ing to be made up of three individuals of about equal size. These are separated as seen on an etched plate by two straight clefts which run through the plate in parallel direc- *Neues Jahrb., 1910, I, 115-117. 96 METEORITES tions. These clefts run parallel to octahedral planes. The clefts divide the etched plate into three nearly equal parts, all of which differ in their etching figures. On two of the parts the usual figures of a fine octahedrite are shown, but they run in different directions, on one as if the section were parallel to an octahedron and on the other as if it were parallel to a hexoctahedron. The third portion ex- hibits at first sight no octahedral figures, but appears like an ataxite. On close examination, however, indications of octhedral figures can be discerned in this portion also. The FIG. 41. Tesselated octahedral figures seen in one of the Bethany meteorites. Some of the lamellae follow the lines of a cube, others those of an octahedron. best explanation of the peculiar structure seems to be that of Berwerth,* who regards the mass as made up of three individuals twinned according to the spinel law, but one of the individuals suffered some subsequent molecular altera- tion. Some iron meteorites on etching do not exhibit the broad, lamellar figures which have been described for octahedral meteorites. Their surface appears, except for occasional inclusions, practically uniform and homogeneous. Yet on close examination their surfaces will be found to be crossed here and there and in different directions by long, straight, narrow, slightly depressed lines. Scattered among these are shorter lines, also running in various directions. Such lines were first observed and described in detail by Neumann *Sitzb. Akad. d. Wiss., Wien., 1902, HI. STRUCTURE OF METEORITES 97 and they are therefore generally known as Neumann lines. Meteorites which exhibit them are also found to possess a cubic cleavage which can best be observed by cutting a section of the meteorite partly in two and then breaking the remainder. The fractured surface then presents an ap- pearance like that of broken galena. The Neumann lines always exhibit some definite relation to the cubic struc- ture, by either running as diagon als of faces or to central points of a cubic edge. The appear- ance of these lines and their relation to a cube are shown in the accom- panying figure (Fig. 42) as drawn by Huntington* from a section of Coahuila. The lines are regarded as representing narrow lamellae more easily dissolved by acid than the intervening portions. They lie in twinning relation to the main indi- vidual. According to Linck| the twinning plane may be the octahedron or both the twinning plane and the growth plane may be the trisoctahedron, 211. Irons which show the Neumann lines also exhibit an oriented sheen. This is FIG. 42. Upper figure shows Neumann lines seen on a section of the Coahuila iron. Lower figure shows the relations of these lines to those of a cube. After Hunt- ington. *Am. Jour. Sci., 3, 32, 284. fZeitschr. fur Kryst., 1892, 20, 209-215. 98 METEORITES believed to be caused by reflection of light (i) from minute, square pits which are negative crystals of a tetrahexahedron and (2) from the Neumann lines. Owing to the various cubic features which characterize irons of this type, they are known as hexahedrites. They form a small but well- marked group. The only member of the group observed to fall is Braunau. Other typical hexahedrites are Coahuila, Hex River, Lick Creek, Murphy, Nenntmannsdorf, and Scottsville. These exhibit an uninterrupted cubic structure and are known as normal hexahedrites. Others which are cubic show a more or less coarse-granular structure and are known as granular hexahedrites. The individual grains on these show Neumann lines which differ in direction. The size and shape of these grains also varies in different meteor- ites. Bingera, Holland's Store, Indian Valley, Summit, and Tombigbee River are examples of this group. While the normal hexahedrites appear alike, they exhibit considerable difference in their resistance to acid. Some, like Fort Duncan, dissolve with difficulty while Lick Creek and Scottsville etch very easily. Other iron meteorites on etching show neither octahedral figures nor Neumann lines, and neither octahedral nor cubic cleavage. Some show a fine-granular structure but others even under the microscope exhibit no division into grains which can be detected. They are, therefore, except for accessory minerals, quite structureless. Peculiar streaks or clouds characterize some, but they are neither octahedral or cubic in their arrangement. To this third group of iron meteorites the name of ataxites is applied, the name mean- ing "without arrangement." Such meteorites, because of their lack of characteristic figures, form a difficult group to distinguish from terrestrial irons. Some of them are regarded by Berwerth as showing traces of octahedral structure and to have been produced by heating of octa- hedral meteorites.* Such an origin for some ataxites is rendered probable by the fact that heating an octahedral iron destroys the octahedral structure, as shown in the accompanying figure (Fig. 43). The upper figure shows a *Sitzb. Wien Akad., 1905, Vol. 114. STRUCTURE OF METEORITES 99 section of Toluca exhibiting the usual etching figures of that meteorite, while the lower figure shows the change made in the appearance of the iron by heating it to a tem- perature of 950 C. for seven hours. The octahedral struc- ture is practically destroyed and an appearance closely resembling that of many of the ataxites is produced. On the other hand not all ataxites can have been produced in this way since some of them have too high a content of nickel to have had octahedral struc- ture. The ataxites fall for the most part as regards composition into two groups, one being high (15 to 20 per cent) in nickel, the other low (5 to 7 per cent). A few are intermediate be- tween these in their nickel content. A study by the writer* of the com- position of the iron meteorites showed that a definite re- lation apparently exists between their composition and structure or that their com- position apparently controls their structure. Thus the hexahedrites all contain about 6 per cent of nickel, the oc- tahedrites from 7 to 15 per cent and one group of the atax- ites a still higher percentage. It appears, therefore, that in cooling from the original magma, a meteoric iron which *Pubs. Field Mus. Geol. Ser., 1907, 3, 106. FIG. 43. Effect of heating a section of the Toluca meteorite for 7 hours at 950 C. Upper figure, before heating. Lower figure, after heating. The heating destroyed the lamellar structure and produced an appearance like that of some ataxites. After Berwerth. 100 METEORITES contains 5 to 7 per cent of nickel will crystallize in the cubic form, one containing between 7 and 15 per cent will crys- tallize in the octahedral form, and one containing a percent- age of nickel greater than this will not crystallize at all. Further, among the octahedral irons the percentage of nickel will influence the width of the bands, the bands being narrower as the percentage of nickel increases. Accessory minerals occur in all the iron meteorites. Their position and form may or may not be independent of the structure of the nickel-iron. Thus in the octahedral irons pyrrhotite may occur in nodules of various shapes and sizes without regular arrangement, or it may occur in plate-like forms arranged parallel to the faces of a cube or dodeca- hedron. In the latter case the term Reichenbach lamellae is applied to the forms. Schreibersite may likewise be ir- regularly distributed or it may occur in plate-like forms arranged according to the planes of a dodecahedron. The latter forms are known as Brezina lamellae. Graphite no- dules also occur in the octahedral irons but they are always irregularly distributed. They may be individual or inter- grown with pyrrhotite or schreibersite. Frequently, inclu- sions of these minerals are surrounded by a border of kama- cite. Within, this follows the outline of the inclusion but on its outer side it is not in accord with the octahedral struc- ture. Cohenite is a common ingredient of the coarse octa- hedrites, usually in the form of prismatic crystals lying in the bands of kamacite. Among the hexahedrites, schrei- bersite in the needle-like form known as rhabditeis a common constituent. It is usually regularly distributed in oriented positions. Schreibersite also frequently occurs in the hex- ahedrites in the form of large inclusions resembling hiero- glyphic characters. Pyrrhotite is not as abundant in the hexahedrites as in the octahedrites but may occur in both oriented and non-oriented positions. Daubreelite is com- mon and characteristic of the hexahedrites, usually occur- ring intergrown with pyrrhotite in parallel plates. Graphite is of rare occurrence in the hexahedrites. Among the ataxites accessory constituents are usually rare and of small size. When they do occur a zone of STRUCTURE OF METEORITES ^j ; \/ 101 slightly different appearance from the rest of {He ^ usually surrounds them in much the same manrieFas a bor- der of kamacite surrounds accessory minerals in the octahe- dral irons. Two or three of the ataxites are rich in rhabdite. Other minerals besides those mentioned which sometimes take part in the structure of iron meteorites are chromite, diamond, amorphous carbon, lawrencite, chrysolite, forste- rite, and quartz. No marked feature so far as known at- tends the distribution of these. Brezina* gives the following as the order of cooling or solidification of the constituents of the iron meteorites: Daubreelite, pyrrhotite, graphite, schreibersite, cohenite, chromite, swathing kamacite, band kamacite, taenite, and plessite. STRUCTURE OF IRON-STONE METEORITES The iron-stone meteorites pass, as has been said, into the iron meteorites on the one hand and the stone meteorites on the other. Yet within their boundaries they present well- marked characteristics of structure. The pallasites, which most nearly resemble the iron meteorites, consist of a sponge- like mass of nickel-iron, the pores of which are filled with chrysolite. The proportion of metal to silicate in pallasites varies in different falls and in individuals of the same fall. Thus in individuals of Brenham part has the pallasite struc- ture and part the octahedrite structure, and while the ma- jority of the individuals of the fall are pallasites, some are entirely octahedrites. In all pallasites the metal shows octahedral figures on etching. The chrysolite element is usually in the form of rounded or angular grains. Some of the grains attain a diameter of a centimeter or more, but a size of about half a centimeter is more common. Often the grains exhibit crystal planes, but a rounding of the solid angles, as if the surface had been fused, usually obscures the crystal forms. The grains are usually surrounded by a band of kamacite which accommodates itself to their form. This shows that the metal solidified subsequent to the si- licate. In the other groups of iron-stone meteorites the sponge-like structure is far less noticeable or if it occurs, *Denkschr. Wien Akad., 1905, 88, 641. 102 METEORITES tnefe is- greater irregularity in the size and shape of the pores in which the silicates occur. The metal tends to ag- gregate in large nodules at times and the silicates do like- wise. Again there may be a uniform dotting of metal as seen on a. section surface, with similar dotting of silicates interspersed. By the gradual diminution of the amount of metal, these iron-stone meteorites of which Mincy and Crab Orchard are good illustrations, pass over to the structure of the stone meteorites. STRUCTURE OF STONE METEORITES CHONDRITIC STRUCTURE A structure peculiar to about 90 per cent of all stone meteorites consists in their being made up of rounded grains orspherules. These grains or spherules are named chondri, (dim. chondrules) from the Greek %6v&po$, a grain, and meteorites largely or partly made up of them are known as chondrites. In size chondri vary from that of a walnut to a dust-like minuteness. The larger number are about the size of millet seeds. The form of chondri is generally spher- oidal, but varies from essentially spherical to mere irregular fragments. Some chondri are flattened or oval and others show apparent deformation subsequent to their origin. In the latter, depressions or projections occur which often look as if a hard chondrus had pressed against another soft one during the process of formation. The deformed chondri pass by every gradation into those which appear to be rock fragments with rounded angles. The surface of the chon- drus is rarely smooth, being usually rough or knobbed. From many friable meteorites individual chondri can easily be isolated, but if the meteorite is at all coherent the chon- dri break with the rest of the mass. The color of chondri is usually white or gray, but some are brown to black. As they are often of the same substance as the groundmass in which they are imbedded they may differ little in color from it. On this account and on account of an ill-defined contour they may be overlooked and a crystal may be considered porphyritic, which is really part of a chondrus. Usually, STRUCTURE OF METEORITES 103 however, the chondri are plainly marked on a polished sec- tion by differences in color and contour. In structure chon- dri may themselves be granular, porphyritic or coarsely or finely fibrous. They may consist of a single crystal indi- vidual, in which case they are said to be monosomatic, or of several individuals, when they are said to be polysomatic. True monosomatic chondri are confined almost exclusively to the mineral chrysolite. They may be known by their simultaneous extinction in polarized light. Polysomatic chondri may be made up of different minerals as well as different individuals and may show more than one kind of structure, i. e., a chondrus may be granular in one portion and fibrous in another. The following minerals are noted by Tschermak as forming chondri, their relative abundance being in the order named: Chrysolite, bronzite, augite, pla- gioclase, glass, and nickel-iron. Chrysolite chondri usually contain large quantities of glass of a dark brown color. This may be arranged (a) in the form of alternate layers, in which case a marked rod-like or lamelliform appearance is produced, (b) forming a base in which the mineral is de- veloped porphyritically, (c) occurring in the center of a crystal, or (d) forming a net-work. Polysomatic chondri of the latter sort are especially liable to be mistaken for those of enstatite since they simulate the fibrous appear- ance of the latter. Occasionally the crystallization may have proceeded only far enough to produce skeletal or branching growths of the mineral among glass. Both mon- osomatic and polysomatic chrysolite chondri may have the arrangement of a well-marked rim about a spherical interior. This rim may, in the polysomatic chondri, be composed of many individuals. Such a rim is often dark from a content of iron and pyrrhotite. Chromite, either in minute grains or in dust-like aggregations, also forms a common inclusion usually near the surface of the chondrus. The quantity of opaque inclusions may be so great as to give the chondrus a black color. Such chondri associated with those of light color are to be found in the stones of Knyahinya, Mezo- Madaras, and others. The constituent minerals of such chondri are chiefly chrysolite and enstatite. Enstatite 104 METEORITES chondri are usually of a finely fibrous character. The fibers instead of radiating from a center as do those of spherulites usually diverge from an eccentric point (Fig. 44). This eccentric arrangement constitutes one of the most marked features of these chondri and separates them sharply from any formation seen in terrestrial rocks. The FIG. 44. Microscopic section of the Homestead meteorite, showing an eccentric radiating enstatite chondrus and a porphyritic and a granular chrysolite chondrus. X65. After Tschermak. enstatite chondri have less glass than those of chrysolite. Monosomatic chondri of enstatite have never been observed, the large crystal individuals showing, as a rule, no tendency to a spherical form. Besides enstatite chondri having an eccentric arrangement of fibers, there occur those which are confusedly fibrous, and these may pass into those which STRUCTURE OF METEORITES 105 have a netted appearance from crossing fibers. Such chon- dri, cut at right angles to the fibers, show the fibers to have a concentric arrangement. The chondri already mentioned, which are granular in part and in part fibrous, are usually made up of the two minerals chrysolite and enstatite. These minerals may be present in about equal quantity or either may be in excess. Usually the enstatite together with glass appears to occupy the intervening spaces between the chrysolite grains, indicating that it is of later formation. Augite chondri are not common but occasionally occur. They often show a structure which indicates repeated twinning. The mineral may appear also in the form of grains, usually of a green color. These grains can be dis- tinguished from chrysolite by their behavior in polarized light. Chondri containing plagioclase in any large quantity are rare but have been observed by Tschermak in the stone of Dhurmsala. The plagioclase alternates in bands with chrysolite and is in excess. .Chondri also occur which are composed almost exclusively of glass, the only indication of the presence of other minerals being in the presence of forked microlites which may be referred to enstatite. Occasionally these microlites are of a pronounced star-like form. Chondri, or at least rounded spheres of nickel-iron, occur in some meteorites, but are not common. All gradations occur from chondri which contain grains of nickel-iron to complete spheres of nickel-iron. In the stone of Renazzo such spheres have a covering of brown glass. Some of the spheres or rounded fragments also contain pyrrhotite, but pyrrhotite of itself has never been seen to form chondri. A more or less complete rim of metal is characteristic of many chondri. The metal may occur in the form of rounded grains or as a continuous periphery. It has been suggested by Daubree that such a rim shows that the chondrus has been subject to the reducing action of hydrogen. Besides the chondri colored black by inclusions of iron and pyrrhotite, previously described, black chondri which consist chiefly of maskelynite or granular plagioclase, occur in the stones of Alfianello, Chateau Renard, and others. These chondri are transparent and colorless about their rim, but in the interior 106 METEORITES are totally black from inclusions of angular or rounded grains, some of which are shown by their brown color to be pyrrhotite. A gathering of grains at the center distinguishes these chondri from those previously described in which the rim was black. Besides complete chondri, fragments repre- senting various portions of a complete chondrus occur. FIG. 45. Microscopic section of the Dhurmsala meteorite, showing a large, somewhat porphyritic chrysolite chondrus enclosing a smaller one. X8. After Tschermak. These may, on account of their shape, be very misleading, as they may be taken for porphyritic individuals or for portions of a foreign stone if their previous chondritic origin is not recognized. Tschermak states that fragments of chondri are most numerous in the stones whose chondri have well-marked contours. So far as the association of STRUCTURE OF METEORITES 107 chondri is concerned it is to be noted that chondri of more than one of the kinds above described usually occur promis- cuously scattered through the same stone. There is no gathering of them into groups according to the minerals they contain. Occasionally one chondrus encloses another (Fig. 45), and still more rarely two may be joined together. FIG^ 46. Microscopic section of the Mezo Madaras meteorite, showing fragments of chondri. Fragments of enstatite, chrysolite, and nickel- iron chondri can be recognized. X?o. After Tschermak. Broken fragments of chondri commonly occur in the stone with complete chondri. Two fragments of the same chon- drus are, however, rarely if ever found in juxtaposition. Hence there must have been considerable separation of the fragments before consolidation of the stone took place, (Fig. 46). 108 METEORITES The conditions which have brought about the formation of chondri are not well understood, though the question has been much discussed and various hypotheses have been sug- gested. The views of earlier observers were to the effect that the chondri represented fragments of pre-existing rock which, by oscillation and consequent attrition had obtained a spherical form. Sorby regarded chondri as produced by cooling and aggregation of minute drops of melted stony matter. Tschermak considers their origin similar to that of the spherules met with in volcanic tuffs which owe their form to prolonged explosive activity in a volcanic throat, breaking up the older rocks and rounding the particles by constant attrition. Different views are, however, held by Brezina, Wads- worth, and others, these believing that the chondri have been produced by rapid and arrested crystallization in a molten mass. Objections to theories of the first class are to be found (i) in the fact that the chondri usually have rough-knobbed surfaces instead of smooth ones, such as attrition might be expected to produce; (2) in the regularly eccentric form of most enstatite chondri, which attrition would be likely to destroy; and (3) in the fact that fragments of a pre-existing rock ought to show the constitution of the rock as a whole instead of a specialized structure. Objections to theories of the second class are to be found chiefly in the clearly frag- mental nature of most chondritic meteorites. It is in their variation from the surrounding ground mass and in the eccentric arrangement of their fibers that chondri differ chiefly from the spherulites of terrestrial rocks. Stone meteorites without chondri, the achondrites, usually differ considerably in structure from the chondrites although various gradations are to be seen. Porphyritic, ophitic and granular structures occur and the resemblance to ter- restrial rocks is much closer than in the chondrites. There are differences, however, in the fact that the granular meteorites are only of fine grain and the ophitic and por- phyritic ones vary in size of grain. Chassigny shows the most typical uniformly granular STRUCTURE OF METEORITES 109 structure, consisting as it does of isometric grains resting near one another. Occasionally angular gaps are filled by weak, doubly refracting, transparent, maskelynite-like substances. Ibbenbiihren and Manegaum, both consisting of enstatite, are similar, although in Ibbenbiihren the grain, according to Tschermak, is not quite uniform since small grains lie be- tween the larger ones. Angra dos Reis is distinguished by a fine-granular structure and is so loose that pieces can be rubbed between the fingers. Lodran also shows, with the exception of the fine iron network, a structure of isometric grains which are numerously bounded by crystal faces com- posed of nickel-iron. Nowo Urei possesses a peculiar struc- ture. Between grains of olivine and augite there lies a fine- grained aggregate consisting of nickel-iron, a graphitic substance, and diamond. It seems to be in the form of a dark network with its meshes filled by silicates since a great number of dark particles are bordered by silicate grains. Some varieties of magnetite-olivinite from Taberg in Sweden show similar structure, the magnetite appearing in the same form as the nickel-iron and carbonaceous sub- stances in Nowo Urei. All eukrites and shergottites show an ophitic structure. This is wont to be better developed the coarser the grain and can usually be recognized macroscopically in represen- tatives of these groups. Anorthite appears in lath-shaped individuals and augite fills the spaces. In Jonzac the ophitic structure is beautifully and uniformly developed and the grain coarse. The plagioclases are 5 mm. in length and sometimes 12 mm. In Stannern and Juvinas the grain is variable in size not only in different stones but in one and the same. This change is so strong in Stannern that Tscher- mak considered the meteorites as consisting of three kinds of stones, and distinguished coarse-granular, radiated, and compact portions. Shergotty shows on the other hand very uniform size of grains. According to Tschermak there oc- cur in the howardites portions with ophitic structure which he regarded as fragments of eukrites. In an essentially uni- form granular structure it is common to find single individ- uals more or less sharply distinguished by their size, also 110 METEORITES partly well-formed crystals and partly fragmental individ- uals. As a rule they are the same minerals as those form- ing the chief mass of the stone, but exceptionally consist of other constituents. Thus in Bustee appear diopside and enstatite; in Shalka bronzite; in Manegaum chrysolite; in Bishopville enstatite and accessory plagioclase; in the how- ardites anorthite, pyroxene, and chrysolite. The ortho- rhombic pyroxenes at times reach the size of a centimeter. Bustee shows an almost porphyritic structure, since the large crystals are very prominent in a fine-grained ground mass. In Shalka the larger individuals are grouped here and there so that the coarser crystalline portions can be seen. The mesosiderites also consist essentially of a uniformly granular aggregate of iron, olivine, enstatite, and to some extent plagioclase in which the olivine often appears por- phyritic and at times in crystals which, according to Reich- enbach, in Hainholz reach a size of \]A. centimeters, and according to Kunz in Mincy 10 centimeters. In the gra- hamites, which are nearly related to the mesosiderites and are distinguished only by the greater quantity of plagio- clase, the structure is variable. According as the plagio- clase or the augite reaches the stronger development, the structure appears either ophitic or granular and porphyritic, the olivines reaching, according to Brezina, a size of iy centimeters, but since in the mesosiderites and grahamites the nickel-iron appears often in the form of chondri, and according toTschermak glass is sometimes present, no typ- ical crystalline granular structure can be said to be present. Also in ophitic structure crystals of augite or anorthite appear often porphyritic. In many achondrites, mesosid- erites and grahamites portions occur which have the ap- pearance of concretionary formations and have a more or less sharp boundary. As a rule only the quantity of the constituent seems to be different from that of the main mass, but at times a different structure may be seen. Thus in Juvinas portions occur without ophitic structure 3 cen- timeters in size, dark and rich in augite and metallic particles. Perhaps here also belong the already mentioned granular, STRUCTURE OF METEORITES 111 hard, fine-grained and easily separable portions of Man- bhoom which are 2 centimeters in size and resemble the howardites. Reichenbach observed in Hainholz a com- pactness of structure toward the peripheral portion of the meteorite. As a rule crystalline granular meteorites possess a compact structure but the howardites are an exception and form the passage from the achondrites to the chondrites. Large cavities in which the constituents show crystals are especially well developed in Juvinas. Such druses fur- nished Rose measurable crystals. Estherville also shows occasionally a drusy structure. Both Reichenbach and Newton observed that the single constituents of the stone meteorites, silicates or nickel-iron, often show a regular ar- rangement when light is reflected from a fractured surface or from polished faces. To see this arrangement requires, of course, careful examination, but with some care parallel systems of lines crossing at right angles may be observed. The lines seldom run straight, usually crooked. They are abundantly interrupted and often return on themselves. Recognition of the lines is the most difficult in the mete- orites rich in chondri and of coarse structure; also if the structure is fine and uniform. Bluff, Crab Orchard, Hes- sle, Pultusk, Renazzo, Siena, Tomhannock Creek, Weston, Vaca Muerta, and Wold Cottage, furnish good examples of these lines. Newton thought that these line systems indi- cated that the same forces that produced octahedral figures in the iron meteorites had controlled the arrangement of the iron particles in the stone meteorites, and compared the structure with that of graphic granite. As a rule the different stones of one and the same fall show in all essential points the same structure. Exceptions how- ever occur. Of the numerous stones which fell at Home- stead and which show the habit of a normal gray chondrite, one differed. This was a compact, dark or grayish green, poor-in-chondri stone. Among 1200 stones of Pultusk in- vestigated by Rath one was free from chondri and poor in metallic constituents. Rath compared its habit with that of Chassigny and stated that it possessed hardly any simi- larity with a chondrite. Brezina distinguished it on the 112 METEORITES ground of its mineralogical composition as amphoterite-like. According to Denza, of the stones which were simultaneous in fall at Motta dei Conti and Villeneuve those falling at the former locality were richer in metallic constituents, more transparent and of finer grain. According to Tschermak, of the stones which fell at Stannern, some of the smaller were compact and homogeneous, or were plainly crystalline and of breccia-like character. CHAPTER XI COMPOSITION OF METEORITES ELEMENTS The following elements have been found in meteorites in amounts sufficient for quantitative determination: Aluminum Iridium Potassium Argon Iron Radium Calcium Magnesium Ruthenium Carbon Manganese Silicon Chlorine Nickel Sodium Chromium Nitrogen Sulphur Cobalt Oxygen Tin Copper Palladium Titanium Helium Phosphorus Vanadium Hydrogen Platinum These occur as follows: Aluminum occurs combined with silica in the stony meteorites, chiefly in feldspars, and perhaps also as a con- stituent of some pyroxenes and chromites. It is much less abundant than in terrestrial crustal rocks. Argon has been found as an included gas. Calcium occurs in stony meteorites as an ingredient of anorthite and pyroxene, also in the sulphide oldhamite. Carbon occurs (i) amorphous, (2) as graphite, (3) as dia- mond, (4) forming carbides of iron, nickel and cobalt and silicon, (5) as a constituent of carbon monoxide, dioxide, and marsh gas, (6) as a constituent of other hydrocarbons, and (7) probably as carbonates. Chlorine is knpwn to occur only in combination with iron to form lawrencite, but other modes of its occurrence are not unlikely. Chromium occurs in combination with iron and sulphur to form daubreelite, with iron and oxygen to form chromite, 113 114 _ METEORITES and probably also in the metallic state alloyed with iron and nickel. Cobalt occurs alloyed with iron and nickel in nickel-iron and takes part with these metals in the composition of carbides, phosphides, oxides, and probably sulphides. Copper occurs in the form of an alloy in nickel-iron, from which it is apparently never absent. Helium has been found as an included gas. Hydrogen occurs as a gas either pure or combined with carbon. It also takes part in the composition of hydro- carbons and perhaps ammoniacal salts. If the water sometimes found in meteorites is of pre-terrestrial origin, this also represents a hydrogen compound. Iridium occurs alloyed with nickel-iron. It is found only in traces. Iron, the most important constituent of meteorites, is chiefly alloyed with nickel, cobalt, and copper. It also com- bines with sulphur, phosphorus, carbon, chlorine and oxygen to form sulphides, phosphides, carbides, chlorides, and oxides. In combination with sulphur and chromium it forms daubreelite and with chromium and oxygen, chromite. It is also an important ingredient of chrysolite and the pyroxenes. Magnesium is next to iron the most important metallic constituent of meteorites. It occurs always in the com- bined form, chiefly as a constituent of chrysolite and the pyroxenes. Manganese occurs in small quantity in the stone meteorites and in traces in the iron meteorites. In the iron meteorites it probably occurs alloyed with the nickel-iron as a metal. In the stone meteorites as MnO it is found both in those portions soluble and those insoluble in HC1, or, in other words, both in chrysolite and the pyroxenes. Its quantity rarely exceeds I per cent. Nickel is like iron a constant and characteristic ingredient of meteorites. As a metal it forms with iron, cobalt, and copper the alloy called nickel-iron which constitutes the larger part of the iron meteorites and is also abundant in stone meteorites. Nickel takes part with iron in the forma- COMPOSITION OF METEORITES 115 tion of phosphides, carbides, and oxides and less prominently of sulphides and chlorides. From the silicates of meteorites it seems to be lacking for the most part, thus presenting a contrast to terrestrial silicates (chrysolite and the pyroxenes) which frequently contain an appreciable quantity. Nitrogen forms a small percentage, usually less than one per cent, of the gases found in meteorites. Its occurrence as an ammoniacal compound in some of the carbonaceous meteorites is also probable. Oxygen occurs chiefly as a constituent of the siliceous minerals of meteorites. It also takes part in the forma- tion of the oxides such as chromite and magnetite found in minor quantities in iron meteorites. It is not found among the gases of meteorites. Palladium has appeared as traces in one or two iron meteorites. Phosphorus occurs chiefly in the form of schreibersite, a phosphide of iron, nickel, and cobalt. It is never lacking from the iron meteorites and is usually found in small quantity in the stone meteorites. Evidence has also been obtained of its occurrence in a free state in one stone mete- orite. Platinum occurs alloyed with nickel-iron. It is found only as traces or a few hundredths of a per cent. Potassium occurs as an ingredient of the feldspars and may also take part in the constitution of some of the py- roxenes. Radium has been found in a single stone meteorite, that of Dhurmsala* in the quantity of 1.12 x io~ 12 per gramme. Two iron meteorites tested at the same time showed none. Ruthenium occurs alloyed with nickel-iron. It is found only in traces. Silicon forms with oxygen and the metals the silicates of which the stony meteorites are chiefly made up. With carbon it forms the rare carbide moissanite, and may be wholly present in this form in the iron meteorites or in part as a metal forming an alloy. Sodium occurs like potassium as an ingredient of the *Strutt, Proc. Roy. Soc., 1906, A, 77, 480. 116 METEORITES feldspars and perhaps also of some of the pyroxenes. It is more abundant than potassium. Sulphur occurs combined with iron, nickel, cobalt, and calcium. It also enters into the composition of a class of hydrocarbons found in meteorites. It is more abundant in the stone than in the iron meteorites but is quite generally present in both. Tin has been reported only in minute quantity and usually in the irons. It is probably alloyed with the nickel-iron. Titanium has often been reported to the extent of a frac- tion of one per cent in the stone meteorites, usually in the insoluble portion and therefore believed probably to occur in the pyroxenes. Of the Angra dos Reis meteorite, which is composed almost wholly of pyroxene, Ti0 2 constitutes 2.39 per cent. Vanadium occurs as traces in the stone meteorites, prob- ably, according to Apjohn, who found it in the Limerick meteorite, as an oxide associated with chromite, this being characteristic of its occurrence in terrestrial rocks. Several other elements have been reported as occurring in meteorites, but the occurrence needs confirmation. Among these are arsenic, antimony, and zinc. Gold was described by Liversidge as occurring in minute yellow grains insoluble in nitric acid in the irons of Boogaldi and Narraburra.* Several elements have been observed in the spectroscopic examination of meteorites which have not been recognized by chemical analysis. Among these are barium and stron- tium, lead and bismuth."); It will be seen from an examination of the list of elements most abundant in meteorites that they are of low atomic weight. Oxygen, silicon, aluminum, magnesium, calcium, sulphur, nickel, and iron are the most abundant elements and all have an atomic weight below 60. Platinum and iridium, the two heaviest elements, occur in but minute quantity. MINERALS The following minerals grouped according to Dana's system have been satisfactorily identified in meteorites: *Jour. Roy. Soc. New South Wales, 1903, 37, 241. fLockyer. The Meteoritic Hypothesis, 1890, 59. COMPOSITION OF METEORITES 117 Diamond C Isometric Elements Graphite C Hexagonal L Nickel-iron Fe, Ni, Co, Cu Isometric Kamacite Fe 14 Ni Isometric Taenite Fe. Ni, Plessite Fe. Ni w Oldhamite CaS Isometric Osbornite Oxysulphide of Sulphides Ca and Ti Phosphides Pyrrhotite FeS Hexagonal and Daubreelite "FeS. Cr 2 S 3 Carbides Schreibersite (Fe, Ni, Co) 3 P Tetragonal Cohenite Fe 3 C Isometric Moissanite SiC Hexagonal Chlorides Lawrencite FeCl 2 Quartz Si0 2 Hexagonal Oxides Tridymite SiO 2 f Hexagonal or } Orthorhombic Magnetite Fe 3 4 Isometric Chromite (Fe, Mg) Cr 2 O 4 Isometric Carbonates Breunnerite (Mg, Fe) CO 3 Rhombohedral 111 1 YYl 1 > a /vl Ol 3 v_yg r-p | lagioclase ^ \\ r\ \ u i (Ca 4 (PO 4 ) 3 Hexagonal Of the above minerals nickel-iron, chrysolite, and the pyroxenes are by far the most abundant. Schreibersite, daubreelite, oldhamite, moissanite, maskelynite, and wein- bergerite are minerals which have not as yet been recognized terrestrially; the others are similar to terrestrial minerals. A fuller account of the above minerals follows. DIAMOND The first discovery of diamond in meteorites was made by two Russian mineralogists, JerofejefF and Satschinoff,* who in 1888 found in the Russian meteorite of Nowo-Urei about I per cent of small, grayish grains whose hardness, specific gravity, chemical composition, and appearance under the microscope all corresponded with those of dia- mond. The remainder of the meteorite was composed of chrysolite, augite, carbonaceous matter, and nickel-iron. Some of the properties of the grains considered as diamond which led to their determination were their insolubility in hydrochloric, sulphuric, and hydrofluoric acids and in aqua regia; their being unaffected by fusion with soda or acid potassium sulphate, and their combustibility in a stream of oxygen. The specific gravity of the grains was between 2.89 and 3.3; hardness greater than that of corundum. By analysis 0.0124 gram of these grains gave: Carbon, 95.40; hydrogen, 3.23; ash, 3.23; total, 101.86. If the estimate that I per cent of the meteorite was diamond was correct, the total amount of diamond in the meteorite was 17.62 grams or 85.43 carats. The grains were of micro- scopic size and no definite crystal forms could be observed. Kunz and Lewisf verified the observations of JerofejefF and *Verh. d. russ. min. Gesell. 1888, 2, 24, 272-292. Also Comptes Rendus 1888, 1 06, 1679-1681. fScience, 1888, u, 118-119. COMPOSITION OF METEORITES 119 Satschinoff to the extent of finding a substance in the Nowo-Urei meteorite which abraded sapphire. The next important discovery of diamond in meteorites was made by Foote and Koenig* in one of the Canyon Diablo irons. In cutting one of these irons for study a cavity was opened which contained small black grains that "cut through polished corundum as easily as a knife through gypsum." These grains were all small and black except one which was white and about }4. mm. ( l / 60 of an inch) in size. This unfortunately was lost in manipulation. The grains were regarded as diamonds because of their hardness and their indifference to chemical reagents. Later Kunz and Hunt- ingtonf by dissolving portions of several Canyon Diablo meteorites obtained white grains having the appearance of beach sand which were unaffected by hydrofluoric or other acids. With these they succeeded in polishing a diamond by the methods usually employed by diamond cutters. Soon after, HuntingtonJ found a vein in one of the Canyon Diablo irons which contained pyrrhotite, silica, and amor- phous carbon and from this he was able to isolate some transparent, colorless diamond crystals showing the forms of octahedrons and hexoctahedrons. About y carat of colorless, yellow, blue, and black diamonds were thus ob- tained by Huntington. Mallard, who also investigated the Canyon Diablo dia- monds, found in a hollow of one of the irons a soft, black, carbonaceous substance in which were round, black grains from ]4. to i mm. in size, which had sufficient hardness to scratch the cleavage surface of a colorless diamond. Friedel| | obtained from one of the irons brownish-gray grains 0.5 to 0.8 mm. in size, resembling carbonado. These had a spe- cific gravity of 3.3 and o.oi 56 grams and yielded on analysis : C = 99.36 Fe 2 Q 3 = 1.28 100.64 *Am. Jour. Sci., 1891, 3, 42, 415-417. fAm. Jour. Sci., 1893, 3, 470-473. JProc. Am. Acad. Sci., 1894, 29, 204-211. Comptes Rendus, 1892, 114, 812-814. ||Bull. Soc. Franc. Min., 1892, 15, 258-263. 120 METEORITES Later, small, transparent, and colorless diamond grains were found by Friedel in Canyon Diablo. Moissan* obtained by solution of a fragment of Canyon Diablo weighing 4 grams, three forms of carbon, (i) dust-like, carbonaceous particles, (2) rounded, compact fragments, and (3) crumpled, thin particles of brownish color. After treatment of this mixture with boiling sulphuric and hydrofluoric acids and potassium chlorate, two yellowish, bort-like fragments were obtained which had the hardness of diamond. Later,| Moissan dissolved a mass of the Canyon Diablo iron weigh- ing 53 kgs. (116 pounds) in hydrochloric acid and obtained about 800 grams of carbonaceous residue. In this he found diamond, both as very small, black, rounded grains, and as transparent, drop-shaped or rounded octahedral forms. Derby} and Cohen both examined specimens of Canyon Diablo for diamonds without success. Both used for the tests complete individuals weighing about 200 grams each which they dissolved in dilute HC1. The residues obtained were completely soluble in stronger acids. These tests show that diamond is not uniformly distributed through the Can- yon Diablo meteorites. Where it occurs it is found to be most abundant near nodules. In Carcote, a crystalline chondrite, Sandberger|| found dull black grains, hardness 9, not affected by acids, which he regarded as weathered carbo- nado. Weinschenk^f found in the residue of Magura which was insoluble in acids, colorless grains and splinters partly isotropic and partly doubly refracting which scratched ruby and gave CO 2 on burning. The presence of diamond in this meteorite was thus indicated. No other meteorites have been reported to contain diamonds. Moissan exam- ined Kendall Co., Dehesa, and Toluca for diamond without success. The Ovifak iron likewise yielded negative results to the investigations of Moissan and Cohen. By seeking *Comptes Rendus, 1893, 116, 218-224 and 288-290. fComptes Rendus, 1904, 139, 773-780. JAm. Jour. Sci., 1895, 3, 49, 108. Meteoreisenstudien, xi, A. N. H. Wien, 1900, 15, 374. ||Neues Jahrb 1889, 2, 180. TIAnn. Wien Mus., 1889, 4, 99-100. Comptes Rendus, 1895, 131, 483-486. COMPOSITION OF METEORITES 121 to reproduce experimentally the conditions under which diamonds seemed to have been formed in the Canyon Di- ablo meteorites, Moissan* was able to produce artificial dia- monds. Moissan's method consisted in strongly compress- ing pure sugar charcoal in a cylinder of soft iron and closing this by a plug of the same metal. This he placed in a crucible containing about 200 grams of molten iron, melted it by means of an electric furnace, withdrew the crucible at once from the furnace and cooled it as rapidly as possible. The object of the sudden cooling was to form a crust on the mass so as to exert a pressure on the interior as the latter cooled, since iron, like water, expands as it solidifies. Water may be employed as a cooling medium but owing to the formation of a badly conducting layer of steam, immer- sion in molten lead for cooling purposes was found to be preferable. After cooling, the iron was dissolved in hydro- chloric acid and a residue consisting of graphite, a maroon- colored variety of carbon, carbonado, and diamond was obtained. This residue was treated with aqua regia, hot sulphuric acid, hydrofluoric acid, potassium chlorate, and fuming nitric acid and the residues then left were treated with liquids of different densities to separate them. From a separation with bromoform, small fragments having the form, hardness, luster, and chemical composition of diamond were obtained. Pure, limpid diamonds in some cases were found which reached a diameter of 0.5 mm. Analogous to this discovery it may be noted that micro- scopic diamonds were found in several hard steels by Rossel.f It may also be noted that the stone meteorites which con- tain diamonds have a composition similar to that of the peridotites in which the South African diamonds are found. Carbon in graphitic cubic form was noted by Haidinger and Partsch in the Magura meteorite{ and regarded by them as a pseudomorph after pyrite, especially as planes believed to be those of the pentagonal dodecahedron were observed. Rose later suggested an origin of the cubes from diamond *Comptes Rendus, 1893, 116, 218-224, and 1894, 118, 320-326. fComptes Rendus, 1897, 123, 113. JPogg. Ann., 1846, 67, 437-439- Abh. Berlin Akad., 1863, 40 and 1872, 532-533. 122 METEORITES and showed by experiment that diamond heated out of con- tact with air becomes opaque and of graphitic appearance. The cubes from Magura were described more fully later by Brezina,* who stated that they reached a size of 2.5 mm. The planes were somewhat arched and the solid angles rounded. Planes of the dodecahedron and tetrakishexahe- dron, the latter having the symbols 310 and 320, were found modifying the crystals. The carbon of which they were composed was partly earthy and grayish-black in color and partly foliated and of shining metallic luster. The scales of the latter variety showed an arrangement parallel to the three axes of a cube. Similar cubes, though smaller, were isolated in large numbers by Fletcherf from the Youndegin meteorite and called by him cliftonite in honor of R. B. Clifton, pro- fessor of physics at Oxford. These were grayish-black, opaque crystals averaging ^4 mm. in thickness, having a predominant cubic form which was modified occasionally by the dodecahedron and a tetrakishexahedron. Rounded and depressed planes were also observed. Some individuals were found to be hollow, others to have a shelly structure. No cleavage was discernible. Hardness was 2.5; specific gravity 2.12; streak black. The chemical characters agreed completely with those of graphite. Fletcher regarded the crystals as a distinct form of carbon deserving the rank of a new species, but the weight of opinion at the present time tends to consider them as pseudomorphs after diamond, for which the name cliftonite can be conveniently retained. Rose's experiments, which as previously remarked showed that diamond can be completely converted to a graphitic form like that of cliftonite by continued heating out of contact with air, makes this origin seem more probable. Rose found that the high temperature of the electric furnace was necessary for the change, the temperature at which cast iron melts having no effect. Cliftonite was also observed by Fletcher in the iron of Cosby Creek. Hunt- ingtonf found cliftonite in Smithville in the forms of *A. N. H. Wien, 1889, 4, 102-106. fMin. Mag., 1887, 7, 124-130. jProc. Am. Acad. Sci., 1894, 29, 255. COMPOSITION OF METEORITES 123 cubo-octahedrons, unmodified cubes, and cubes truncated by the dodecahedron and a very obtuse tetrakishexahedron. He also found a skeleton octahedron of graphite ^ of an inch in diameter in a nodule of graphite from Cosby Creek. Cohen and Weinschenk* found cliftonite in Toluca in the form of elongated groups composed mostly of cubes but occasionally containing octahedrons. The largest crystals reached a size of only o.i mm. They oxidized somewhat more slowly to graphitic oxide by treatment with potassium chlorate and nitric acid than graphite from the same meteor- ite. With the exception of Toluca, cliftonite seems to be confined to the coarse octahedral irons. GRAPHITE This substance occurs in grains of sufficient size for ready examination only in the meteoric irons. In these it is usually in the form of nodules but sometimes occurs in plates or grains. The nodules often reach considerable size. One nodule taken from the Cosby Creek iron is as large as a small pear and weighs 92 grams. Even larger ones were found in the Magura iron. Toluca, Cranbourne, Chulafmnee and Mazapil are other irons which contain considerable graphite. Graphite has been estimated to form 1.17 per cent of the mass of Magura and 0.8 per cent of the Cosby Creek iron. The mineral is usually associated with iron sulphide. With this it may be intimately inter- grown or the one may enclose the other. Its texture is compact rather than foliated. Smith found that the mete- oritic graphite oxidized much more rapidly than terrestrial graphite on treatment with nitric acid and chlorate of potash. This feature distinguishes it from the amorphous carbon separated from cast iron. The meteoritic graphite is also very pure. Although occurring in nodules of the size described, which must have segregated from the surrounding mass, the ash amounted, in an analysis made by Smith, to only i per cent. By ether was extracted a small quantity of a substance made up of sulphur and a hydro-carbon, which constituted the only other impurity. Emphasizing *Meteoreisenstudien, A. N. H. Wien, 1891, 6, 140-141. 124 METEORITES the differences between meteoritic and terrestrial graphite Smith was inclined to believe that the graphite of meteorites must have been formed by the action of bi-sulphide of car- bon upon incandescent iron rather than that it was analo- gous in its origin to terrestrial graphite. Ansdell and Dewar, however, concluded from elaborate comparisons of meteoritic and terrestrial graphite that they were similar in origin, and were formed by the action of water, gases, and other agents on metal carbides. > AMORPHOUS CARBON Meteorites of the group known as carbonaceous meteor- ites, as well as some others, are permeated by a dull-black, pulverulent coloring matter which is usually left as a residue on treatment of the meteorite with acid. This residue sometimes amounts to from 2 to 4.5 per cent of the mass. A residue similar in character though smaller in amount is likewise found after dissolving many of the iron meteorites. These residues on being heated in air, glow, usually become lighter in color and give off carbon dioxide. They must therefore be considered practically pure carbon. Berzelius and Wohler believed this carbon to have origi- nated, so far as the carbonaceous meteorites are concerned, from the decomposition of the hydrocarbons of the latter. In this respect they regarded it analogous to terrestrial humus, though of very different origin. Smith considered it similar in origin to the graphite of iron meteorites and Weinschenk believes it similar to one of the forms of carbon produced in the making of cast iron. No indications that it had an organic origin have ever been discovered. NICKEL-IRON Nickel-iron is the substance of which the metallic portion of meteorites is chiefly composed. The iron meteorites consist of it almost wholly and from the stone meteorites it is perhaps never altogether absent, although Roda, Chassigny, Shalka, and Angra dos Reis have been described as without it. In the carbonaceous meteorites it is not present as such but their oxidation products indicate that COMPOSITION OF METEORITES 125 it occurred in them. From this almost universal presence of nickel-iron in meteorites, Bombicci has argued that the magnetism of meteorites is the property by virtue of which they are drawn to the earth, the latter acting as a magnet to attract them. In composition the nickel-iron of meteorites is not a sub- stance of fixed proportions but an alloy of iron and nickel in which the percentage of nickel lies between 6 and 20 per cent, and for the most part below II per cent. In the stone meteorites the percentage of nickel is sometimes higher. Thus in the nickel-iron of Honolulu, Mordvinovka, Nerft, and Middlesbrough percentages of nickel of 37.73, 21.16, 20.94, an d 23.01 per cent, respectively, have been reported. Accompanying the nickel of nickel-iron, cobalt and cop- per seem to be universally present. The percentages of cobalt vary as a rule between 0.5 and 2.5 per cent (in Uri- coechea 2.56 per cent). Those of copper range from traces to a few tenths of one per cent. The percentages of cobalt or copper seem to hold no definite relation to the amount of nickel, although irons rich in nickel are usually corre- spondingly rich in cobalt. The color of nickel-iron varies from iron-gray or steel- gray in alloys poor in nickel, to tin-white and silver-white in those rich in nickel. Under the microscope in reflected light, nickel-iron exhibits a bluish reflection. Nickel-iron is more or less easily soluble in the common cold, dilute acids, also in solutions of copper sulphate, copper chloride, copper ammonium chloride, mercurous chloride, bromine water, and iodine with potassium iodide. By cold, dilute hydrochloric acid (i HC1 : 20 aq.) the nickel-poor alloys are completely and the nickel-rich partly dissolved. The specific gravity of nickel-iron varies chiefly between 7.6 and 7.9, although determinations as low as 6.5 and as high as 8.1 have been reported. Normally it should be higher the greater the percentage of nickel since while the specific gravity of pure iron is 7.88 that of pure nickel is 8.8. In cohesive properties nickel-iron varies considerably, being now hard, now soft, now tensile, now brittle, now malleable, and now non-malleable. Of 52 iron meteorites 126 METEORITES accounts of which were collected by Cohen, 48 were reported to be malleable and 4 not malleable. The malleability of much nickel-iron is attested by the fact that it has been manufactured both by barbarous and civilized peoples into utensils and ornaments such as knives, spearheads, horse- shoes, nails, and rings. In boring and cutting iron meteor- ites very different qualities are exhibited by different indi- viduals, some yielding easily to tools and others only with difficulty. These differences may be due to variations in the quality of the nickel-iron itself or more often probably to the presence of harder minerals, such as cohenite and diamond. The iron of Canyon Diablo shows great resist- ance to tools, due undoubtedly to included diamond. All nickel-iron seems to take a good polish. It is also magnetic and many iron meteorites show polarity acquired probably by induction from the earth. The location of the magnetic poles has been determined near the ends of individuals of Staunton, Welland, Tonganoxie, Bingera, and Imilac. The nickel-iron of meteorites is as a rule quite compact. Fletcher describes cavities bounded by planes (negative crystals) in the meteorite of Greenbrier County, and in Lick Creek portions possessed a porous character. Of the iron meteorites nickel-iron forms the entire substance without regular boundaries except as octahedral or cubic cleavage may appear. In the iron-stone meteorites it may appear either as a network the meshes of which are filled with silicates (pallasites), or as apparently rounded grains united by threads, or as branching threads filling spaces between the silicates (mesosiderites). In the chondritic meteorites nickel-iron takes the form of isolated grains or variously shaped, often toothed flakes filling the spaces between the silicates. Regular forms more or less resembling crystals are occasionally observed. Siemaschko described crystals from Tabory weighing 0.2 grams showing the combina- tion 100, in, no, and hko. Incomplete cubes with vicinal faces of a tetrakishexahedron have been described from Barbotan by Partsch and Pfahler. Goalpara furnished cubelike crystals according to Tschermak, and Tomatlan COMPOSITION OF METEORITES 127 octahedrons according to Shepard. Wohler described six and four-sided forms from Parnallee which he interpreted as fragments of a dodecahedron. Brezina noted crystals of nickel-iron in the druses of Estherville. Besides crystals nickel-iron occurs in rounded forms resembling chondri. Such forms observed in Hainholz reached a diameter of 22 mm. (i inch) and in Mincy 6 cm. (2^/2 inches). Of other chondri nickel-iron often constitutes a large part. In these it takes the form of flakes, foliae, grains, and cuboidal forms having at times a concentric arrangement. Such chondri have been noted in Renazzo. Meso-Madarasz, Borkut, Dhurmsala, Gopalpur, and Tie- schitz. In many of the siliceous chondri of stone meteor- ites nickel-iron often forms a periphery either as separate grains or as a thin, coherent, irregular cover. In Parnallee a cylinder of nickel-iron was observed of the dimensions I x y^ mm. In several of the stone meteorites films and scales of nickel-iron occur, appearing in section as fine metallic veins. Although the nickel-iron of meteorites appears in a pol- ished piece to be a homogeneous substance of uniform composition, investigation shows that it is in reality a com- plex substance, made up of alloys containing different quantities of nickel. The existence and character of these alloys is easily made evident by subjecting a polished sur- face of the nickel-iron to the action of heat, acids, or other etching agent. Figures of a more or less banded character then appear on the surface of the iron showing its complex structure. The discovery of this means of investigating the character of nickel-iron was made, as has previously been mentioned, by Alois von Widmanstatten of Vienna in 1808. The production of these figures by heating can be accomplished by placing a thin section of the meteorite upon an asbestos plate and placing it over a Bunsen burner. According to the degree of oxidation the different alloys then appear in different colors, as for instance blue, purple, and yellow. Although this was the method first employed by Widmanstatten it is rarely used at the present time since the employment of liquid etching agents is simpler and 128 METEORITES gives more delicate results. Of these agents the most con- venient and satisfactory is usually nitric acid. For pre- liminary testing the acid diluted to about one-tenth its normal strength may be applied to a small, flat, freshly filed surface of the nickel-iron. In four or five minutes the character of the figures will usually be roughly outlined. For etching of a plate for careful study of the figures more pains should be taken. Meteorites differ in the degree and speed with which they are attacked, some etching easily and quickly with weak acid, others only after longer treat- ment with stronger acid. The surface of the meteorite to be investigated should be flat and smooth and the larger the surface the greater will be the opportunity afforded to study the details of its structure. Foote Mineral Company of Philadelphia, who have had excellent success in etching meteorites, have given the writer the following details of their method of etching: 1. Wash the specimen with benzine. 2. Lacquer the unpolished back and edges with a lacquer known as "steel gloss," diluting it about one-half with ben- zine. When this side is dry, carefully remove with benzine any lacquer which may have run over the edges onto the polished surface. An electric fan greatly hastens the drying of the lacquer. 3. Lacquer any nodules. They should be completely covered, as they are readily attacked by the acid, and will stain the etched surface. 4. Place the iron so that the polished surface is horizontal. Wash with a 5 to 15 per cent solution of C. P. nitric acid for from 15 seconds to 4 or 5 minutes, until the etching is brilliant. If etched much longer, the iron will darken. When the surface begins to get rough, the maximum bril- liancy has been reached. The acid should be kept as thick and as even as possible by rubbing the plate with a large brush. As the acid becomes discolored, it should be brushed off and fresh acid added. 5. To clean and facilitate rapid drying, quickly put the section into clean warm water (120 to 130 F.) for several minutes, rubbing with a brush. COMPOSITION OF METEORITES 129 6. Dry in a few seconds with blotting-paper. 7. Thickly lacquer the etched surface at once. To avoid oxidizing, the operations from 4 to 7 should be accom- plished as quickly as is practicable, by having all materials at hand. Where possible the writer has found the etching to be more delicately performed if the plate to be etched is dipped into the acid with the side to be etched down instead of up and instead of pouring the acid on the plate. Such dipping facilitates removal by gravity of the products of etching. Nevertheless, in many instances the size and shape of the plate prevent such immersion and the acid must be poured on. Other etching agents besides nitric acid which may be employed are dilute hydrochloric acid, the addition to which of a small volume of choride of antimony is said to lessen subsequent rusting, a solution of sulphate of copper, the deposited copper being removed by ammonia, solutions of chloride of mercury, chloride of gold, chloride of platinum, fused alkalies or bromine water. The figures obtained with some of these agents are said to differ from those obtained in other ways. Upon the great majority of iron meteorites the figures which appear upon etching show the nickel-iron to be made up of three different alloys differing in form, color, luster, and degree of solubility. One of these alloys appears as bands of iron-gray color and dull luster, which on heating are more thickly covered with oxide or on etching are more depressed than the other alloys. The bands cross each other in manifold fashion, and while rather uniform in width in any single meteorite in different meteorites show varia- tions in width from X to 2 mm. and in length from a few millimeters up to 10 cm. (2^ inches). This alloy was called by Reichenbach Balkeneisen or Kamazit from /ea/^af, a pole or shaft, and is known in English as kamacite. Bordering the bands of kamacite appear others which are narrower, silver-white in color and more brilliant in luster. Their substance is less attacked by acids or oxidizing agents and hence they stand out in relief. To this alloy the name Bandeisen or Taenit from raivLa, a ribbon, was given by 130 METEORITES Reichenbach. These two alloys run parallel to and adjoin- ing each other and together form what is known as a lamella. The crossing of these lamellae in network fashion leaves angular spaces or meshes which are often filled by a third alloy generally of darker color and duller luster than the kamacite. This third alloy is known as plessite from ReichenbacrTs name Plessit or Fiilleisen. Its degree of ox- idation and solubility is intermediate between that of kam- acite and taenite. The three alloys together are known as the trias or triad. Meteorites containing or made up of nickel-iron which exhibit these three alloys are known as octahedral meteor- ites or octahedrites since the arrangement of the lamellae in such meteorites proves to be parallel to the planes of an octahedron. Two other classes of iron meteorites as already noted display no such compound structure. These are the hexahedrites or cubic meteorites and ataxites or meteorites without structure. The hexahedrites are made up of but a single one of the above alloys, kamacite, while the ataxites have a diverse composition. KAMACITE Balkeneisen Kamacite is the predominant constituent of nickel-iron. Of the cubic iron meteorites and some of the ataxites it forms practically the entire mass and in the octahedral meteorites it is more abundant than any other constituent. In color it is iron-gray as contrasted to the tin-white of taenite and the usually darker gray of plessite. It is solu- ble in dilute acids of the stronger class such as HC1 1:20 and very slowly even in acetic acid. Its specific gravity ranges from 7.78 to 7.87 and its hardness is between 4 and 5. According to structure, three different kinds of kamacite are recognized: hatched, spotted, and granular. The hatched kamacite (Brezina's schraffirten Kamazit, Reichen- bach's Feilhiebe) is characterized by being covered by networks of fine lines tending to cross at right angles. These are like the Neumann lines of the cubic meteorites on a smaller scale. The spotted kamacite (Brezina's fleckig COMPOSITION OF METEORITES 131 Kamazit) shows varying dark and light spots from unequal reflections of light. The appearance is caused by groups of lines or pits. The spots are of irregular outline and rarely exceed I mm. in diameter. The granular kamacite (Bre- zina's kornige or abgekornt Kamazit) consists of grains separated by rather deep channels. This separation usually appears only after strong etching. The grains may be coarse (i to 2 mm. diameter) or fine (o.i to I mm. diameter.) All these kinds of kamacite may be found in different meteorites. The cubic meteorites are composed almost wholly of hatched kamacite and some of the ataxites almost wholly of granular kamacite. The other kinds of kamacite are seen in the bands of different octahedral meteorites. Several kinds of kamacite are also distinguished according to their position or form. These kinds are known as swath- ing kamacite, swollen kamacite, grouped kamacite, and un- equally grouped kamacite. Swathing kamacite (Brezina's Wickelkamazite, Reichenbach's Fiilleisen or Wulsteisen) is seen enclosing accessory constituents in meteoric irons. It usually forms a band two or three millimeters broad around accessory minerals, following their outlines within and not conforming to the general structure of the meteorite without. Swollen kamacite (Brezina's wulstiger Kamazit) is a characteristic form assumed by the kamacite bands of many of the octahedral meteorites. Such bands are short, swollen in the middle, and often have rounded ends bounded by taenite. Grouped kamacite (Brezina's gescharter Kama- zit) consists of bands lying close together, parallel and gen- erally elongated. It characterizes many octahedral meteor- ites. Unequally grouped kamacite (Brezina's ungleich gescharter Kamazit) consists of grouped bands of different lengths, of which the middle ones are usually the longer. Such bands may be seen in many octahedral meteorites. In addition to these forms of kamacite, certain jagged and angular fragments found remaining behind after the solution in dilute HC1 of the kamacite of many octahedral meteorites prove on analysis to have a composition near that of kamacite. Their lower solubility is believed to be generally due to included schreibersite or cohenite. 132 METEORITES All the above-named forms of kamacite have a chemical composition closely approximating that represented by the formula Fe H Ni of which the percentages are Fe 93.11 per cent, Ni+Co = 6.89 per cent= 100. Analyses of kamacite of the various kinds mentioned are given below, the kamacite of octahedral irons being given first, as it was in these that kamacite was first distinguished. ANALYSES OF KAMACITE I. Kamacite of octahedral meteorites. Fe Ni Co Cu C Total Fe : Ni + Co i 93-Qi 6.22 0.77 tr. 100 13.98:1 2 93-09 6.69 0.25 0.02 100.05 14.09:1 REFERENCES 1. Bendego. Isolated by Derby, analyzed by Florence and Dafert: Ann. Mus. Rio de Janeiro, 1896, ix, 140 and 183. Calculated to 100 after deducting insoluble residue. 2. Welland. Davison: Am. Jour. Sci., 1891 (3), xlii, 64. Plates i to 2 mm. thick, of the color of cast-iron, with wrinkled surface and covered with a thin layer of magnetite; brittle; conchoidal fracture. II. Swathing kamacite. Fe Ni Co Total Fe : Ni+Co 92.62 6.55 0.83 loo- 13.19:1 REFERENCE Glorieta. Cohen and Weinschenk: A. N. H. Wien, 1891, vi, 158. III. Jagged, residual kamacite. Fe Ni Co Cu C Total Fe : Ni+Co i 92.62 6,81 0.57 100 13.19 i 2 93-01 6.25 0.74 100 13.98 i 3 93-27 6.04 0.64 .... 0.05 100 14.67 i 4 93-89 5.30 0.61 .... 0.20 100 16.69 i 5 94-05 5.26 0.57 .... 0.12 loo 16.95 i 6 94-09 5.51 .... 0.05 0.34 100 17.77 i REFERENCES 1. Canyon Diablo. Florence: Am. Jour. Sci., 1895 (3), xlix, 104. Calculated to IOO after deducting 0.31 per cent taenite and 0.35 per cent schrcibersite. 2. Magura. Sjostrom: A. N. H. Wien, 1898, xiii, 484. Calculated to 100 after deducting 0.58 per cent schreibersite. 3. Magura. Manteuffel: A. N. H. Wien, 1892, vii, 156. 4. Staunton. Manteuffel: A. N. H. Wien, 1892, vii, 157. 5. Toluca. Manteuffel: A. N. H. Wien, 1892, vii, 157. 6. Canyon Diablo. Florence: Am. Jour. Sci., 1895 (3), xlix, 104. Calculated to loo after deducting 0.31 per cent taenite and 0.35 per cent schreibersite. COMPOSITION OF METEORITES 133 IV. Angular, residual kamacite. Fe Ni Co Total Fe : Ni+C. 92.94 6.18 0.88 100 13.83:1 REFERENCE Magura. Cohen and Weinschenk: A. N. H. Wien, 1891, vi, 152. TAENITE Bandeisen, Meteorin, Edmondsonite. Taenite is the ingredient of octahedral nickel-irons which occurs in thin plates. These are usually of a nickel-white color though they become by oxidation golden to isabel- yellow. They border the kamacite bands and containing more nickel are less attacked by etching agents. They, therefore, stand in relief. The thickness of the taenite plates may vary from 0.03 to 0.25 mm. Taenite is less liable to decomposition than kamacite, and hence often remains in the form of bright, more or less elastic plates after the decomposition of the mass of a meteorite. These plates" often have a crumpled, wavy appearance. They resemble schreibersite, for which they have sometimes been mistaken, in being strongly magnetic but fuse with more difficulty B.B. Taenite is attacked slowly by cold, dilute acids and con- siderably but not entirely dissolved. Concentrated nitric and hydrochloric acids and copper-ammonium chloride dissolve it completely. Analyses of taenite 'show per- centages of nickel varying from about 13 to 48 per cent, cobalt being also usually reported in quantity up to 2 per cent. These analyses indicate that taenite has not a uniform composition. S. W. J. Smith* has observed that taenite isolated mechanically usually contains less nickel than that isolated chemically through the prolonged action of dilute acid, and states that this indicates that taenite contains considerable kamacite which is dissolved out by the acid. The analyses given below do not bear out this statement however since the percentage of nickel seems to be entirely independent of the manner in which the mate- rial for analysis was obtained. *Phil. Trans. London, 1908, Ser. A., vol. 208, p. 21. 134 METEORITES Furthermore the structure of the taenite bands indicates a complex composition. Tschermak found the taenite lamellae of Ilimae to consist of a fine network of different bodies which he regarded as chiefly nickel-iron mixed with pure iron. Taenite occurs only in the octahedral irons and more abundantly in the fine octahedrites than in the coarse. Thus Cohen* estimated the percentage of taenite in the coarse octahedrite of Wichita as 2.64 per cent, while in the medium octahedrites Toluca and Misteca he regarded it as 6.79 per cent and 6.75 per cent, respectively, and in the fine octahedrites Chupaderos and Glorieta Mountain 10.24 P er cent an d n-35 P er cent. The composition of taenite as shown by various analyses is as follows : ANALYSES OF TAENITE Fe Ni+Co Fe Ni Co Cu C Total + Cu i 86 4.4. 1*1 O2 O Z IOO oo 6 9 2 8c T'T oo j. j . \s& 14 .OO qq oo 6 4" '3-- ^j ... 85 oo A if. v -"- f 15 .OO V :/ IOO oo 6 . o 4. 83 28 16.68 .... 0.04 IOO oo c . 2 . . . wj ... 80 30 19.60 99 90 J 4 . I 6. ... 74 78 24.32 0.33 0.50 99 93 3 . 2 7 - 73 IO 23-63 2.10 1.17 IOO 00 3 . 8 - 73 o 27.00 100 00 2 . 8 9 ... 72 12 27-73 . 02 .... 0. 12 IOO 00 2 7 10 ... 71 29 26.73 1.68 .... 0.30 IOO 00 2 .6 ii ... 70 H 29.74 99 88 2 . 5 12 ... 69 30 29 73 o . 60 .... 0-37 IOO 00 2 4 13 ... 68 1*3 30.85 0.69 0.33 .... IOO.OO 2 . 2 14 ... 65 54 32-87 1.59 .... IOO oo 2 . O IS ... 65 39 33-20 1.41 IOO oo 2 . O 16 ...65 26 34 34 o . 40 .... IOO oo 2 . 17 ...63 55 34 65 i. oi 0.30 o 49 IOO oo 9 18 ... 63 04 35-53 1.43 .... tr. IOO 00 . 8 10 ... 61 89 36.95 0.36 .... 0.80 IOO 00 7 * 7 20 ... 61 87 38.13 .... tr. IOO 00 7 21 ... 57 18 34 oo 0-55 91 73 7 22 . . . . 50 73 47.80 0.63 0.37 0.47 IOO .00 i REFERENCES i. Cosby Creek. Reichenbach, Jr.: Pogg. Ann., 1861, xciv, 258. Plates 8 cm. long and 2^ cm. broad, mechanically isolated by Reichenbach, Sr. Mean of three analyses. *Meteoreisenstudien, ii, A. N. H. Wien, 1892, vii, 160, 161. COMPOSITION OF METEORITES 135 2. Charcas. Meunier: Ann. Chim. et Phys., 1869 (4), xvii, 31. Particles mechanically isolated by their color. . 3. Caille. Meunier: Ann. Chim. et Phys., 1869 (4), xvii, 32. Net-like web, isolated by means of dilute nitric acid. 4. Casas Grandes. Tassin: Proc. U. S. Nat. Mus., 1902, xxv, 73. 5. Kenton Co. Nichols: Pub. Field Col. Mus., 1902, Geol. Ser., i, 315. Thin, tin-white, elastic magnetic plates, 4 mm. square, with finely ribbed surface. 6. Welland. Davison: Am. Jour. Sci., 1891 (3), xlii, 66. Mechanically isolated plates jVVo mrn - thick, silver-white to bronze-yellow, flexible and elastic. 7. Staunton. Cohen and Weinschenk: Ann. Wien Naturhist. Mus., 1891, vi, 146. Gray, relatively thick and brittle plates. Isolated by dilute HC1. Cal- culated to 100 after deducting schreibersite. 8. Cosby Creek. Smith: Comptes Rendu, 1881, xcii, 843. Little thin plates of white metallic color left after dissolving the iron in acid. ' 9. Canyon Diablo. Tassin: Smithsonian Misc. Coll., 1907, i, 212. Calcu- lated to 100 after deducting 0.26 per cent schreibersite. 10. Magura. Weinschenk: Ann. Wien Naturhist. Mus., 1889, iv, 97. Thin, tough, silver-white lamellae soluble with difficulty in acids. Isolated by dilute HC1. Calculated to 100 after deduction of schreibersite. 11. Cranbourne. Flight: Phil. Trans. London, 1882, No. 171, 888. White, flexible, magnetic, triangular or rhombic mechanically isolated plates. 12. Misteca. Cohen: Ann. Wien Naturhist. Mus., 1892, vii, 152. Dull and brittle plates. Isolated by HC1. Calculated to 100 after deducting schreibersite. 13. Canyon Diablo. Florence: Am. Jour. Sci., 1895 (3), xlix, 105. Thin tin-white, flexible plates. Calculated to 100 after deduction of 3.60 per cent schreibersite. 14. Wichita Co. Cohen and Weinschenk: Ann. Wien Naturhist. Mus., 1891, vi, 155. Isolated by dilute HC1. Calculated to 100 after deduction of schreiber- site. 15. Chupaderos. ManteufFel: Ann. Wien Naturhist. Mus., 1892, vii, 150. Brittle, tin-white plates. Isolated by HC1. Calculated to 100 after deducting schreibersite. 16. Toluca. Cohen and Weinschenk: Ann. Wien Naturhist. Mus., 1891, vi, 137. Tin-white, flexible plates. Isolated by HC1. Calculated to 100 after deducting schreibersite. 17. Canyon Diablo. Fahrenhorst: Ann. Wien Naturhist. Mus., 1900, xv, 376. Thin, flexible plates partly appearing made up of many lamellae, light-yellow or grayish. Schreibersite, 2.34 per cent deducted. 18. Glorieta Mountain. Cohen and Weinschenk: Ann. Wien Naturhist. Mus., 1891, vi, 137. Tin-white, flexible, grouped plates. Isolated by HC1. Calculated to 100 after deducting schreibersite. 19. Bischtiibe. Cohen: Ann. Wien Naturhist. Mus., 1897, xii, 54. Large, flexible plates with included schreibersite. Isolated by HC1. 20. Penkarring Rock. Fletcher: Min. Mag., 1899, xii, 174. Thin, flexible plates. Analysis calculated to 100 after deducting 4.18 per cent schreibersite. 21. Medwedewa. Berzelius: Pogg. Ann., 1833, xxxiii, 133. Analysis of skeleton material left behind after dissolving in HC1. 22. Beaconsfield. Sjostrom: Monatsberichte Berlin Akad., 1897, 1041. Tinto silver-white, lustrous plates. Iron determined by difference. The analyses, as will be observed, show variations of composition from Fe 7 Ni to FeNi. While this variation is a wide one it is evident that it is between certain limits, and that it would be incorrect to ascribe too indefinite a com- position to taenite. 136 METEORITES PLESSITE Fulleisen The term plessite or fulleisen was applied by Reichenbach to the nickel-iron alloy filling the spaces or fields between the lamellae of octahedral meteorites, and the term is still used in this sense. This alloy is usually of darker color and duller luster than kamacite, and as a rule is surrounded by a border of taenite. The amount of plessite occurring in this way varies in different irons. When the fields are more abundant than the lamelFae, plessite is correspondingly more abundant, but in some irons there is an almost complete absence of fields and therefore of plessite. In some of the finely laminated octahedrites, as Butler, there are no dis- tinguishable fields but plessite forms the principal mass of the meteorite with octahedral lamellae scattered through it either singly or in bundles. In several pallasites also pies- site strongly predominates. While plessite thus appears to have a certain individuality, a little examination of it with a lens or even the naked eye, especially as it occurs in the fields of octahedral meteorites, shows that it is not a homo- geneous substance. Often the fields show a structure which simply repeats that of the rest of the meteorite on a smaller scale. The plessite is in these cases made up of kamacite and taenite perhaps less perfectly separated than in the larger lamellae. Davison* carefully separated by hand- picking the two constituents of this character making up the fields of the Welland meteorite and found that they cor- responded both physically and in chemical composition to the kamacite and taenite of the same meteorite. The kam- acitic portion of the plessite was made up of minute rods T /6 to T / 3 mm. in diameter and the taenitic portion of bands I /i3o to J /22o mm. thick. It is evident from all these consid- erations that much plessite represents simply less individ- ualized portions of the meteorites in which it occurs and its composition would therefore correspond in general to that of the meteorite itself. Its composition may therefore be safely considered as intermediate between that of kamacite *Am. Jour. Sci., 1891, 3, 42, 65. COMPOSITION OF METEORITES 137 and taenite, or between Fei 4 Ni and FeNi. From analogies with the behavior of other alloys it has been suggested by Fletcher* that the behavior of these nickel-iron alloys may be illustrated by that of a fused mixture of silver and cop- per. When the percentage by weight of silver is 72, and that of copper 28, says Fletcher, solidification begins, not at a temperature between 960 and 1083, the solidifying temperatures of silver and copper, respectively, but at a temperature below both, namely 770. The solid which first separates has the same percentage composition as the original mixture; the part still fused has thus itself the same percentage composition as before, and continues to solidify at the same temperature, and in the same way, until the solidification is complete. Such a mixture, having a definite composition and a definite temperature of solidification, was for a time regarded as a definite chemical compound with a complex chemical formula, but on microscopic examina- tion the resultant solid was found to be heterogeneous; minute particles of the silver and copper were seen to lie side by side, the particles being granular or lamellar in form according to the circumstances of the cooling. If the per- centage of silver is different from 72, whether it be higher or lower, the solidification begins at a higher temperature than 770; whence the mixture containing 72 per cent of sil- ver has been conveniently termed eutectic (i.e., very fusible). When the silver is in excess of 72 per cent, the excess of silver gradually collects together and solidifies at various parts of the cooling fused mass; the still fused portion thus gradually becomes poorer in that metal, and the tempera- ture, instead of remaining constant, gradually falls during the separation of the solid. At length the percentage of silver in the fused portion falls to 72 per cent and the tem- perature to 770; the solid which now begins to form is no longer pure silver, but a material containing 72 per cent of that metal; and it continues to have the same percentage composition as the surrounding liquid, and the temperature of solid and liquid to be 770, until the solidification is com- *Fletcher, Introduction to the Study of Meteorites. British Museum, Nat. Hist., London, 1909, p. 40. 138 METEORITES plete. The final solid thus consists of blebs of silver scat- tered through a fine groundmass of eutectic mixture of silver and copper. Similarly, if the copper is in excess of 28 per cent, the final solid consists of blebs of copper scat- tered through a fine groundmass of eutectic mixture of silver and copper. Hence Fletcher suggests that plessite is a eutectic of nickel-iron, and that the kamacite and taenite first separated from it. Prof. Rinne* is of the opinion, however, that separation took place after solidification. Hence he proposes the term eutropic instead of eutectic as it avoids the conception of fusion. It is evident that further study of these points is desirable. Experimental investigation should show whether iron-nickel alloys have a eutectic and if so the composition of that eutectic. According to the different development of taenite in plessite the appearance of plessite may vary from a dull, dark-gray to a lighter-colored, glittering alloy. In some meteorites also, especially the pallasites, plessite may have a spotted appearance like spotted kamacite. OLDHAMITE This is a simple calcium sulphide which has been found in a few meteorites. It is light brown in color and trans- parent when pure. Hardness 4, specific gravity 2.58. It is soluble in water, isotropic, and has equal cleavage in three directions, hence is doubtless isometric. It was first discovered by Maskelyne in the Bustee meteorite, occurring in rounded grains coated with gypsum through alteration. He gave it the name oldhamite in honor of R. D. Oldham, Director of the India Geol. Survey. The mineral was found occurring in chestnut-brown spherules scattered at one end of the meteorite. These spherules, as nearly as can be determined from Flight's figuref range from 6 mm. in diameter down. The composition of the spherules was determined by Maskelyne, in two analyses, I and II, to be the following: *Neues Jahrb. fur Min., 1905, Bd. I, 122. fA Chapter in the History of Meteorites, 1887, 118. COMPOSITION OF METEORITES 139 I II Oldh 't / Calcium monosulphide 89.369 90.244 \ Magnesium monosulphide 3 . 246 3 . 264 Gypsum 3-951 4. 189 Calcium carbonate 3 .434 Trolite 2 . 303 100. 100. Oldhamite was observed in the Hvittis meteorite by Borgstrom* as present in the form of transparent grains 3 mm. in diameter and of a brownish yellow color. The min- eral was isotropic and easily dissolved, with evolution of H 2 S, by acetic acid. Little honey-yellow grains seen in Bishopville were regarded by Maskelyne as oldhamite and Brezina saw similar ones in Aubres. As the water extract of Morristown contained calcium sulphide, Merrillf con- cluded that oldhamite was present in that meteorite. Tas- sin concluded^ from an analysis of the non-magnetic portion of Allegan that oldhamite occurred in that meteorite also, but could obtain no visible evidence of its presence. Mer- rill also observed the mineral in Indarch. Oldhamite is notable in being a meteoritic mineral which has not yet been discovered terrestrially. Vogt observed it in furnace slags. OSBORNITE In the brown spherules of oldhamite found in the Bustee meteorite, Maskelyne observed minute octahedrons having a golden color and. luster. 1 1 They were not affected by acids nor by fusion with potassium carbonate. Ignited in dry chlorine they glowed, lost their metallic luster and left a deliquescent residue. Only .002 of a grain could be obtained for analysis, hence only qualitative determinations could be made. These showed the presence of calcium, sulphur, and an element which was regarded as either titanium or zir- conium. On account of the resistance to acids which the mineral exhibited, Maskelyne thought it should probably *Die Meteoriten von Hvittis and Marjalahti, Helsingfors, 1903, 35. fAm. Jour. Sci., 1896, 4, 152-153. |Proc. U. S. Nat. Mus., 1908, 34, 433-434. Arch. Math. Nat., 1890, 14, 72. ||Phil. Trans., 1870, 109, 198-202. 140 METEORITES be regarded as an oxysulphide of calcium and titanium or zir- conium rather than a simple sulphide. He named the min- eral osbornite in honor of Mr. Osborne, who had brought the Bustee meteorite to England and collected information regarding the fall. The mineral has never been reported from any other meteorite. PYRRHOTITE Troilite Iron sulphide was early recognized as a constituent of meteorites, Count Bournon having noted it in several meteorites in 1802. He regarded it, however, as pyrite. The easy decomposability by acids of the meteoritic iron sulphide as compared with pyrite was soon noted, however, and the sulphide was then assumed to be pyrrhotite, especial- ly since Rose in 1825 found crystals in Juvinas which gave forms identical with those of pyrrhotite. Several later investigators found the composition of the iron sulphide of iron meteorites to be FeS, thus differing from that of pyrrhotite which was regarded as FenSi 2 . Accordingly Haidingerin 1863* proposed the name troilite for the simple iron sulphide of meteorites, a name given in honor of the Jesuit priest, Dominico Troili, who had described in 1766 the Albareto meteorite and mentioned the occurrence of iron sulphide in it. Inasmuch as Rose's measurements indi- cated that the iron sulphide of one of the stone meteorites was pyrrhotite while analysis gave the composition of the iron sulphide of iron meteorites as FeS, it was for some time thought that the iron sulphide of stone meteorites should be regarded as pyrrhotite and that of iron meteorites as troilite. This view was adopted by Cohenf in his earlier work but modified later. MeunierJ was of the opinion that all mete- oritic iron sulphide was pyrrhotite, and adopted for pyr- rhotite the formula Fen Si 2 . Hintze considers troilite a *Sitzb. Wien Akad., xlvii, II, 287-289. fMeteoritenkunde, 1894, Heft I, 190. JMeteorites, Paris, 1884, 62. Handbuch der Mineralogie, 630. COMPOSITION OF METEORITES 141 synonym for pyrrhotite but gives pyrrhotite the formula FeS. Analyses show no essential difference between the iron sulphide of iron and stone meteorites and according to Busz* the measurements upon which Rose based his deter- mination of the crystal form of the Juvinas troilite were approximate only and showed great variations. The latest work upon the subject has been that of Allen and associates, f who, although they do not seem to have worked upon meteoritic iron sulphide directly, concluded from an elaborate general study of the sulphides of iron that troilite should not be considered to differ mineralogic- ally from pyrrhotite, it being simply the end point of a series of solid solutions. The larger percentage of iron in the pyrrhotite of meteoritic irons as compared with that in terrestrial pyrrhotite they regard as due to the excess of iron present at the time of formation of the sulphide. It is thought probable by them that the pyrrhotite of stone me- teorites does not contain this excess of iron. It may be de- sirable, therefore, to use the general name pyrrhotite for the iron sulphide of meteorites in general, and that plan is here adopted. It may be remarked that troilite differs from terrestrial pyrrhotite in being easily decomposed by acids, in leaving no residue of sulphur after such decompo- sition, in being as a rule non-magnetic, and in having a higher specific gravity than terrestrial pyrrhotite. The color of meteoritic pyrrhotite varies from bronze- yellow to tomback brown, the darker color tending to be acquired on exposure. Hardness 4, specific gravity 4.68- 4.82. Streak grayish black. Luster metallic. Fuses easily B.B. to a black, magnetic globule. Easily decomposed by hydrochloric acid with evolution of hydrogen sulphide. Crystal form as determined by Rose in Juvinas (Fig. 47) and Brezina in Bolson de Mapimi, hexagonal. The follow- ing forms are reported: o (oooi), r (1010), t (1120), s (loll), P (2021), v (1121) Brezina reports cleavage II to the base and the pyramidal * Neues Jahrbuch, 1895, i, 125-6. fAm. Jour. Sci., 1912, 4, 33, 213. 142 METEORITES planes P (2021) striated II to the base. Linck at first concluded from the cleavage of troilite that its form was isometric, but later* regarded the form as like that of pyr- rhotite. Such crystals as have been found have been of small size and with imperfect, rounded faces. They occur in druses of the stone meteorites. Be- sides Juvinas they have been reported from Richmond, Farmington, and Estherville. Most of the pyrrhotite in stone meteorites occurs in the form of small, scattered grains without angular or regular outline. These grains rarely reach a diameter of more than a few r VK ~7; L ^" r . millimeters although Rose reported one tic. 47. Pyrrhotite . ^ .. . ... ,, . crystal from the in Gruneberg 1 3 mm. in diameter. Vein- Juvinas meteorite. u 'k e masses r C m. long have also been After Rose. -i i T i i described. In the chondntic meteorites pyrrhotite is especially common in scattered grains. It is frequently intergrown with nickel-iron in all proportions but it also occurs singly. The darker color and dull luster of pyrrhotite easily distinguish it from nickel-iron and further distinction may be obtained by allowing a meteorite section to stand a few minutes in a solution of copper sulphate. Copper will then be deposited upon the nickel- iron but not upon the pyrrhotite, since pyrrhotite does not reduce copper from a solution of copper sulphate. So far as has been observed, the distribution of the pyrrhotite in the chondritic meteorites bears no especial relation to the structure of the chondri. In the iron meteorites pyrrhotite is common and abundant. It especially characterizes the medium octahedrites, and occurs in them in much larger masses than in the stone meteorites. A nodule of pyrrhotite isolated by Smith from Cosby Creek weighed 200 grams, and one obtained from Magura was 13 centimeters long. Nodules of a spheroidal form are a common mode of occurrence. Other common forms are cylindrical, lens-shaped, oval, and indented. In certain sections a ring-like form may be presented. A re- *Cohen, Meteoritenkunde, Heft I, 1894, 191, and Heft II, 1903, 248. COMPOSITION OF METEORITES 143 markable mode of occurrence is that forming the so-called Reichenbach lamellae (Fig. 48). These are small plates of pyrrhotite distributed through the nickel-iron of some mete- orites and oriented according to the planes of a cube. These lamellae though occasionally larger, range as a rule from 0.1-0.2 mm. in thickness and from 1-3 cm. in length. Staun- 1 >' FIG. 48. Reichenbach lamellae as seen in the Ilimae meteorite. ton, Trenton, Victoria West, Cleveland, Merceditas, and especially Jewell Hill are meteorites which exhibit these lamellae. The lamellae were first observed by Reichen- bach in Lenarto and Caille. Tschermak noted their orientation parallel to the three planes of a cube in the Ilimae and other meteorites,* and Brezina proposed the * Denkschrift Wien Akad., 1871, 31, 192-4. 144 METEORITES term Reichenbach lamellae* by which they are now generally known. The formation of Reichenbach lamellae in a meteorite must occur prior to that of the nickel- iron. Schreibersite sometimes borders the lamellae. Associated with large nodules of pyrrhotite other min- erals are frequently found, often with a more or less zonal arrangement. Graphite and schreibersite thus frequently FIG. 49. An iron meteorite of the Canyon Diablo fall which is perforated probably by the fusing out of a nodule of pyrrhotite. The weight of this individual is 219 Ibs. join with pyrrhotite and occasionally chromite and daubreelite. A common zonal arrangement, seen especial- ly in Wichita and Canyon Diablo, is an interior of pyr- rhotite surrounded by a layer of graphite and that by one of schreibersite or cohenite. ' Pyrrhotite, like impurities in artificial iron, tends to be most abundant toward the periphery of a meteoric in- dividual. Owing to this fact and its easy fusibility it may play an important part in the shaping of the pittings which characterize the surface of meteorites. More striking still are the results it produces when an entire nodule fuses out *Denkschrift Wien Akad., 1880, 43, 13-16. COMPOSITION OF METEORITES 145 and leaves a hole which may pass entirely through an iron meteorite. (Fig. 49.) Most of the analyses of meteoritic pyrrhotite show a close, approximation to the formula FeS the percentages of which are: Fe 63.60, S 36.40. In the 26 analyses here given only three depart far from this formula and these standing alone can hardly be regarded conclusive. As a rule the quantity of nickel, cobalt, or copper present is very small. Meteor- itic pyrrhotite in this respect therefore differs to a marked degree from schreibersite, since in the latter nickel and co- balt are essential constituents. ANALYSES OF PYRRHOTITE i Fe 6q 28 Ni Co Cu Si0 2 2 64. IQ O I 3 T. 6^ 03 4 5 63.84 63 82 tr 6 8 9 10 ii 12 13 63.80 63.61 63.53 63.48 63.47 63.41 63-40 63 . 3; 0.42 . 20 .... o . 08 H. J J3 63 . 34 :i 17 18 10 63.28 63.00 62.65 62.38 62.32 0-45 tr. I . 02 .... I .96 0.32 I C8 .... o . 67 tr. 0.56 20 . > 62.21 o 16 o 56 21 22 23 24 11 62.01 61.80 61 . ii 59-01 58-94 58.07 0.89 1.56 0.14 .... 0.42 4-34 1-52 tr Fe+Ni+Co S Total +Cu : S 34-72 100 0.929. 35-68 100 0.969 36.07 100 0.985 36.16 100 o . 989 37-36 101 .18 i . 023 36.28 100.08 0-993 36.33 TOO. 12 0.997 36-05 IOO 0.986 36.21 99.69 0.996 35-89 IOO 0.988 36.29 99.70 i 36.21 99.81 0-993 35-91 99.26 0.990 36.66 IOO I .Oil 35-59 99-59 0.976 35-27 99.96 o . 954 35-39 IOO 0.958 35-67 99-01 0.994 36. 10 IOO 0.988 35-05 98.49 0.977 38.28 101.18 . 064 36.64 IOO .012 39 56 100.67 H3 40.03. 99.18 . 118 39-99 99-35 . 118 36.07 IOO 0.990 " REFERENCES 1. Bendego. Dafert: Ann. Mus. Nac. Rio de Janeiro, 1896, ix, 129. After separation of 5.26 per cent of insoluble residue, consisting chiefly of daubreelite and schreibersite. Traces of nickel, cobalt and silica were found. The material analyzed consisted of non-magnetic particles which dissolved without evolution of sulphur in moderately concentrated HC1. 2. Tennasilm. Schilling: Archiv. fur die Naturk. Liv., Esth. u. Kurlands, 1882 (i), ix, .109. Calculated to 100 after deducting 0.357 per cent residue. The lens showed an admixture of nickel-iron. 146 METEORITES 3. Rowton. Flight: Phil. Tr. London, 1882, No. 171, 896. Iron determined by difference. 4. Sokobanja. Losanitsch: Ber. der deutsche Chem. Gesell. Berlin, 1878, xi, 97. Copper not present. Sulphur determined by difference. 5. Nenntmannsdorf. Geinitz: Neues Jahrbuch, 1876, 609-610. 6. Cosby Creek. Smith: C. R., 1875, Ixxxi, 978. 7. Cranbourne. Flight: Phil. Tr. London, 1882, No. 171, 891. Mean of 4 analyses after deduction of 0.215 P er cent an( J 0.297 P er cent insoluble residue, 0.13 per cent chlorine and 0.207 per cent sulphur. 8. Bear Creek. Smith: Am. Jour. Sci., 1867 (2), xliii, 66. Calculated to 100 after deducting 1.81 per cent of residue. 9. Cosby Creek. Smith: C. R., 1875, Ixxxi, 978. 10. Seelasgen. Rammelsberg: Monatsber. Berlin Akad., 1864, 368. Sulphur determined by difference, 0.64 per cent manganese. 11. Jelica. Losanitsch: Berichte der deutsche Chem. Gesell. Berlin, 1892, xxv, 880. No nickel or cobalt. 12. Casas Grandes. Tassin: Proc. U. S. Nat. Mus., 1902, xxv, 72. Color of material analyzed brass to bronze yellow; weakly magnetic. 13. Seelasgen. Rammelsberg: Monatsber. Berlin Akad., 1864, 368. 14. Sierra de Chaco. Domeyko: C. R., 1864, Iviii, 555. 15. Bjurbole. Borgstrom: Bull. Com. Geol. de Finlande, 1902, Nr. 12, 25. Mean of several closely agreeing analyses. 16. Steinbach. Winkler: Nova acte Halle Akademie, 1878, xl, No. 8, 357. 17. Cosby Creek. Rammelsberg: Pogg. Ann., 1864, cxxi, 367. Calculated to loo after deducting 0.74 per cent residue. Sulphur determined by difference. 18. Tazewell. Smith: Am. Jour. Sci., 1855 (2), xix, 156. 19. Seelasgen. Rammelsberg: Zeit. der deutsch. geol. Gesell. Berlin, 1870, xxii, 894. Sulphur determined by difference. Calculated to 100 after deducting 0.18 per cent P. 20. Cosby Creek. Smith: Am. Jour. Sci., 1876 (3), xi, 433. Dissolved by HNOs from a nodule of graphite; 0.30 per cent MgO. 21. Nocoleche. Cooksey: Records of Australian Museum, Sydney, 1897, iii, 53. The material analyzed was treated with mercuric chloride for some time to remove 34.6 per cent intermingled nickel-iron. 22. Cosby Creek. Rammelsberg: Pogg. Ann., 1864, cxxi, 367. Calculate* loo after deducting 0.60 per cent residue. Sulphur determined by difference. 23. Danville. Smith: Am. Jour. Sci., 1870 (2), xlix, 91. 24. Toluca. Meunier: Ann. Chim. et Phys., 1869 (4), xvii, 42. 25. Jelica. Meunier: Ann. geol. de la Pen. Balkanique, 1893, iv, 5-10. 26. Beaconsfield. Sjostrom: Sitzb. Berlin Akad., 1897, 1044. Material, separated with great care, consisted of non-magnetic grains ^-2 mm. in diameter. Graphite 0.33 per cent, traces P and Cl. DAUBREELITE This mineral, originally discovered by Smith in one of the Coahuila irons* is an iron-chromium sulphide peculiar to meteorites. Its composition is FeS, Cr 2 S 3 . It is found in nearly all the cubic iron meteorites and has also been identified in theirons of Toluca, Nelson County, Cranbourne, Canyon Diablo and others. It has never been found in stone meteorites. It usually accompanies pyrrhotite, either bor- dering nodules or crossing them in veins. Sometimes, how- *Am. Jour. Sci., 1876, 3, 12, 109; 1878, 3, 16, 270. COMPOSITION OF METEORITES 147 ever, it occurs as thin plates or grains. It is black in color, has a black streak, is of metallic luster, brittle and not mag- netic. It is infusible before the blowpipe and becomes mag- netic in the reducing flame. It is not attacked by hot or cold hydrochloric acid, but is completely dissolved by nitric acid without the separation of free sulphur. This solubility distinguishes it from chromite. Its system of crystalliza- tion is not known though it exhibits rectangular and tri- angular partings which indicate one of the systems of high symmetry. Sp. gr. =5.01. Meunier obtained the mineral artificially by treating an alloy of iron and chromium at a red heat with hydrogen sulphide. Smith obtained as the mean of three analyses: S 42.69, Cr 35.91, Fe 20.10=98.70 The theoretical composition is: S 44.3, Cr 36.3, Fe 19.4 = 100 The chromium content of iron meteorites soluble in nitric acid or aqua regia has usually been ascribed to daubreelite but as Cohen points out* this is only allowable when suffi- cient sulphur is present to form this mineral. SCHREIBERSITE Phosphornickeleisen, Rhabdite, Dyslytite, Lamprite, Glanzeisen, Partschite The name schreibersite was first applied to a nickel-iron phosphide occurring in the form of large crystals or grains. To occurrences of the same substance in needle or plate-like forms the name rhabdite was long applied as it was at first thought to be a different mineral from schreibersite. The composition and physical properties of rhabdite were, however, shown by Cohen to be the same as those of schrei- bersite and rhabdite is now regarded as a variety of schreiber- site. The term nickel-iron phosphide is sometimes used as a general one to include both schreibersite and rhabdite, but in these pages the term schreibersite is used as a general name for the species. The color of schreibersite on fresh fracture is tin-white; *Meteoritenkunde, Heft II, 256. 148 METEORITES it tarnishes, however, readily to bronze-yellow. When fresh its lighter color distinguishes it readily from pyrrhotite, kamacite and plessite but not so readily from taenite and cohenite. In fact it has often probably been confounded with taenite but though resembling that alloy in color it can readily be distinguished from it by its lack of elasticity. Schreibersite is extremely brittle. Its hardness is about 6.5. Determinations of its specific gravity vary from 6.3 to 7.3, but the majority of determinations give a value near 7. Schreibersite is strongly magnetic and may be made to acquire and hold polarity. It occurs in the form of crystals, grains, foliae, and especially as needles, to which latter form the name rhabdite (pa/35o5, a rod) is applied. The habit of the crystals varies from compressed-prismatic to vertical-tabular. The crystals sometimes reach a length of 5X inches (14 cm.) as in Carlton. Owing to the round- ing of planes and angles crystallographic determinations have not thus far been possible. Often a hollowing of the ends of the crystals such as is characteristic of pyromorphite occurs and in one such hollow on a crystal from Toluca, taenite was found. Cleavage in three perpendicular direc- tions is usually observable, that perpendicular to the length of the crystal being the best defined. The tetragonal sys- tem of crystallization is thus indicated. By allowing a crystal of Schreibersite to fall on paper, separation into cuboidal forms often takes place. Disintegration also often occurs by mere standing, thus suggesting that a condition of tension like that shown by Prince Rupert drops, exists. In the form of grains or flakes Schreibersite may be dis- cerned in an etched section or its presence may only become known upon dissolving the nickel-iron. In Ilimae isolated grains were observed by Tschermak to be collected in the neighborhood of Reichenbach lamellae. Often Schreibersite is intergrown with pyrrhotite and graphite. Graphite no- dules may serve as a nucleus for the growth of Schreibersite crystals or they may be surrounded by an envelope of the latter mineral. The needle-like or lath-shaped forms of Schreibersite known as rhabdite usually have angular termi- nations. They are usually very thin in proportion to their COMPOSITION OF METEORITES 149 length. In thickness Cohen obtained measurements vary- ing from .001 to .05 mm. while in length some were 5 mm. long. In the latter direction, however, the measurement might well be incomplete since the needles break easily. The width of the 5 mm. crystal was 1.5 mm. Angular measurements made by Deecke and Scheerer showed values close to 90. Thus the tetragonal form is further indicated. Vertical striations often characterize the needles. In the irons of Santa Rosa, Seelasgen, Braunau, and Misteca the arrangement of rhabdite was found to be parallel to cubic faces. In Indian Valley, Kunz and Wein- schenk noted an arrangement parallel to the Neumann lines. In Hex River the rhabdite is arranged in parallel zones, but the needles are differently oriented in each. Zones poor in rhabdite or free from it lie between those rich in rhabdite. There is also a. difference in the size of the needles in the different zones, and the smaller needles appear more numer- ous and more crowded than the larger. A similar arrange- ment was observed by Brezina in the iron of Holland's Store. In Braunau, according to Tschermak, there are gradations from true rhabdite needles to schreibersite-like foliae, the latter being bounded by three planes perpendicular to each other and arranged partly parallel to the faces of the main individual and partly parallel to the twinning lamellae. This is explained by Tschermak as a simultaneous crystal- lization of nickel-iron and schreibersite, the slow-forming twinning lamellae giving the schreibersite more opportu- nity to extend in the direction of breadth. Somewhat similar foliae were isolated by Cohen from Hex River. The largest of these was 2.6 mm. long, 1.6 mm. broad, but most were much smaller. Still there was complete separa- tion in form and dimensions between needles and plates. The plates were usually bounded by six planes at right angles to each other, the four subordinate ones being prob- ably cleavage planes. On some of the edges of the planes angles of about 150 could be noted. The surface of the planes was either smooth or corrugated, the corrugations running sometimes in two directions. 150 METEORITES Brezina* noted an arrangement of schreibersite lamellae parallel to the planes of the dodecahedron. These he found in Tazewell, Ballinoo, and Narraburra. The plates have the same relation to schreibersite that those of pyrrhotite (Reichenbach lamellae) have to that mineral except that the pyrrhotite lamellae are arranged parallel to the cube, while those of schreibersite are parallel to the dodecahedron. FIG. 50. Brezina's lamellae (dodecahedral schreibersite lamellae) as seen in the Narraburra meteorite. Cohen has given the name of Brezina's lamellae to schrei- bersite plates arranged in this way. The accompanying figure (Fig. 50) shows the Brezina lamellae in Narraburra as observed by Liversidge.f Numerous analyses of schreibersite and rhabdite have been made all indicating a formula (Fe, Ni, Co) 3 P. The ratio of Fe : Ni+Co varies, but considered as 3 : I which seems to be about the average, the percentages become Fe 62.6, Ni+Co 21.9, P 15.5. In view of the large, constant percentage of nickel, Borgstrom has urgedf that the formula *Sitzb. Wien. Akad., 1904, 113, 1-7. tjour. Roy. Soc. N. S. W., 1903, xxxvii, PI. xvii. JDie Met. v. Hvittis u. Marj., Helsingfors, 1903, 67. COMPOSITION OF METEORITES 151 of schreibersite should be regarded as Fe 2 Ni P, in which case the percentages would be Fe 55.5, Ni 29.1, P 15.4. The percentage of Ni+Co is much less variable apparently in rhabdite than in schreibersite. A complete list of analyses follows. In cold, dilute acids and in copper ammonium chloride schreibersite is insoluble, the latter property fur- nishing a means of distinguishing it from cohenite and taenite. Unlike these minerals also it does not reduce cop- per from copper sulphate solution. By warm, concentrated HC1 or aqua regia schreibersite is easily dissolved and thin plates are attacked by cold, dilute HC1. For isolation of schreibersite Meunier recommends boil- ing a powder of the mineral with a concentrated solution of copper sulphate, separation of precipitated copper with fuming nitric acid, treatment of the residue with a magnet to separate it from graphite, and solution of any pyrrhotite present by treatment with dilute nitric acid. Before the blowpipe schreibersite fuses easily to a magnetic globule. After boiling with nitric acid, ammonium molybdate gives a yellow precipitate of ammonio-phospho-molybdate. Heat- ing of the fine powder with magnesium wire in a closed tube and treatment of the assay with water gives hydrogen phosphide, H 3 P, recognizable by its garlic-like odor. Schreibersite is almost universally present in the iron meteorites, in fact is perhaps never lacking from them. The distribution of schreibersite in an individual meteorite is usually quite irregular. Thus from three different pieces of Glorieta Mountain Cohen obtained values varying from 2.85 to 8. 1 1 per cent of schreibersite according to what part the sample was taken from, and similar determinations of Toluca by different authorities show percentages varying from 0.34 to 4.93 per cent. It is quite impracticable, therefore, to determine accurately the amount of schreiber- site in a meteorite by calculation from the amount of phos- phorus obtained in a single analysis. As between schreibersite and rhabdite the octahedrites usually contain more schreibersite, the hexahedrites and ataxites more rhabdite. Yet both may occur in about equal quantity as in Seelasgen, while in Magura and Sarepta 152 METEORITES certain parts contain an excess of schreibersite and others of rhabdite. The iron-stone meteorites usually contain schreibersite, as do probably also the stone meteorites. At least it is customary to assign to this mineral percentages of phosphorus shown by analyses. Yet this amount is sometimes so large, e.g., 0.76-0.91 per cent in the nickel- irons of Nerft, Honolulu, Zsadany and Bachmut and 2.03 per cent in the nickel-iron of Carcote that it is suggested by Cohen* that it is questionable whether this is properly referable to schreibersite. In a later note in view of the discovery of free phosphorus in Saline by Farrington, Cohen suggests that the phosphorus may be present in the free state. As was early pointed out by Smith, schreibersite is a min- eral peculiar to meteorites and one of the most significant in interpreting their origin. Terrestrially phosphides do not occur, since free oxygen changes them to phosphates. The existence of schreibersite in meteorites is therefore, proof of absence of free oxygen where they were formed. Tornebohm has reported schreibersite to be present in the terrestrial iron of Ovifak, his determinations being based on the presence of magnetic particles with metallic luster which did not precipitate copper from a copper sulphate solution and were only slightly attacked by hydrochloric acid. Such a determination is, however, too incomplete to be reliable. In the Santa Catarina iron Daubree found a prismatic crystal terminating in eight planes which he regarded as schreibersite. Also a substance isolated by Derby from this iron by means of dilute hydrochloric acid gave percentages on analysis by Cohen closely approximat- ing the composition of schreibersite. A number of iron or nickel phosphides resembling schreibersite have been obtained artificially. Thus Sidot by allowing vapor of phosphorus to pass over piano wire in a porcelain tube at a red heat and later heating the product, obtained in the interior of the metallic mass hard, steel-colored, four-sided prisms, reaching a centimeter in size and having the com- position Fe 4 P (P 1 2. i per cent). By fusion of calcium *Meteoritenkunde, Heft I, 136. COMPOSITION OF METEORITES 153 phosphide, powdered charcoal, and nickel oxide, Gamier obtained a compound having the formula Ni 5 P. It was in the form of long, prismatic, light yellow crystals determined by Jannetaz to be tetragonal. Their hardness was 5.5; sp. gr. 7.283. By slight heating of finely divided iron in vapor of phos- phorus Hvoslef obtained a compound having the formula Fe 2 P from which by strong heating under a cover of borax, a regulus of dark iron-gray color, brittle, magnetic, and at- tacked neither by hydrochloric or nitric acids resulted. It contained 16 per cent P, corresponding to the formula Fe 3 P. G. = 6.28. Rhabdite-like, steel-gray, brittle, magnetic, tetragonal prisms were found by Mallard accompanying augite and anorthite among the products of a furnace at Commentry. After subtracting Fe and As the composition of the substance corresponded to the formula Fe 7 P 2 . ANALYSES OF SCHREIBERSITE P Fe Ni Co Cu Total Substance Fe+Ni+Co:P taken 1 16.10 72.62 10.72 0.56 loo 0.5375 2.869:1 2 16.04 69.55 H-4 1 loo 2.877 3a 15.74 69.54 13-81 1.31 100.40 0.4499 2.955 3b 15.80 70.07 14.57 0.43 0.03 100.90 0.8030 2.959 4 15.70 61.78 21.93 0.38 0.21 loo 0.6886 2.937 5 15.70 54.12 29.71 0.47 loo 0.5045 2.925 6 15.68 50.52 33.90 0.62 0.22 100.94 0.6761 2.955 7 15.49 63.36 19.63 1.23 .... 99.71 0.4086 2.978 8 15.47 65.75 i8.35 0.43 .... loo.oo 0.5644 2.996 9 15.45 54.43 29.36 0.67 0.34 100.25 0.3260 2.990 10 15.45 7 I -7 12.58 0.32 100.05 0.6490 3.014 11 15.37 58.54 26.08 0.05 tr. 100.04 3.009 12 15.38 63.97 19.15 1.68 100.18 0.4115 3.020 13 15-31 57-46 25.78 1.32 .... 99.87 0.1328 3.015 14 15.01 57-11 28.35 tr 100.47 3.106 15 15.00 64.69 20.11 99.80 3.099 16 14.93 55.15 29.15 0.21 .... 99.44 0.5152 3.086 17 14.88 66.92 18.16 0.62 .... 100.58 0.4023 3.158 18 14.86 56.53 28.02 0.28 99.69 3.113 .... 100.00 0.619 3 J 7 2 0-13 98.97 3 171 .... 100.00 3 . 277 : loo.oo 0.028 3 . 324 ; 99-94 3-419: 13.51 56.12 29.18 98.81 3.443: 154 METEORITES REFERENCES 1. Zacatecas. Scherer: Meteoritenkunde, Heft I, 131. Calculated to 100, after deduction of 4.60 per cent chromite and 0.88 per cent daubreelite. 2. Cranbourne. Flight: Ph. Tr., 1882, 892; mean of two analyses. Nickel determined by difference. "Large, brass-yellow prisms, with distinctly basal cleavage, slowly soluble in muriatic or nitric acid." 33. S. Juliao de Moreira. Cohen: N. J., 1889, i, 220; mean of two analyses; brittle, steel-gray, crystalline fragments, shading to bronze-yellow. 3b. S. Juliao de Moreira. Fahrenhorst: Meteoreisenstudien, xi, A. N. H. Wien, 1900, xv, 389. The determination of Cu was made on 19.183 grams. 4. Kendall County. Scherer: Meteoritenkunde, Heft II, 233. 5. Mount Joy. Fahrenhorst: Meteoreisenstudien, xi, A. N. H. Wien, 1900, xv, 388. Slight mixture of rhabdite. Calculated to 100 after deduction of 0.42 per cent chromite and silicate. 6. Magura. Fahrenhorst: Meteoreisenstudien, xi, A. N. H. Wien, 1900, xv, 3 77. 7. Glorieta Mountain. Cohen and Weinschenk, A. N. H. Wien., 1891, 157; large and very brittle crystals. 8. Bischtiibe. Cohen: Meteoritenkunde, Heft I, 131. Calculated to 100, after deduction of 0.07 per cent residue. 9. Cosby Creek. Fahrenhorst: Meteoreisenstudien, xi, A. N. H. Wien, 1900, xv, 372. 10. De Sotoville. Cohen: Meteoritenkunde, Heft II, 233. 11. Canyon Diablo. Tassin: Smithsonian Misc. Coll., 1908, 50, 212. Flat- tened and angular nodules and rounded grains. 0-7.20. 12. Toluca. Cohen and Weinschenk: A. N. H. Wien, 1891, vi, 138; very brittle crystals, up to 5 mm. in length, tin-white and very lustrous; cobalt prob- ably estimated too high on the basis of the nickel; no copper. 13. Hraschina. Cohen and Weinschenk: A. N. H. Wien, 1891, vi, 149. Fragments. 14. Toluca. Meunier: A. Ch. P. (Paris), 1869 (4), xvii, 45, 57; microscopic scales, slowly soluble in warm muriatic acid; trace of magnesia. 15. Casas Grandes. Tassin: Proc. U. S. Nat. Mus., 1902, xxv, 73. 16. Marjalahti. Borgstrom: Die Meteoriten von Hvittis und Marjalahti, 1903, 66. 17. Beaconsfield. Sjostrom: Monatsberichte Berlin Akad., 1897, 1040. Large crystals. 18. Tazewell. Smith: Am. Jour. Sci., 1885 (2), xix, 157; yellow, irregular spangles. 19. Bohumilitz. Berzelius: Pogg. Ann., xxvii, 131; gilt scales; calculated to 100, after deduction of 2.04 Si. and 1.42 C. 20. Canyon Diablo. Florence: Am. Jour. Sci , 1895 (3), 49, 107. From a vein enclosed by cohenite Tin found qualitatively. 21. Cambria. Silliman and Hunt: Am. Jour. Sci, 1846 (2), ii, 375; blackish- gray folia, mixed with bright flakes; calculated to IOO after deduction of 10 per cent Si. The analysis gave a sum of only 90 per cent. 22. Elbogen. Berzelius: Pogg. Ann., 1834, xxxiii, 137. 23. Canyon Diablo. Tassin: Smithsonian Misc. Coll., 1908,50,211. Broad, thin, dark steel-gray, flexible, magnetic lamellae. = 7.09. 24. Cranbourne. Flight: Ph. Tr., 1882, 892; very brittle, coarse powder, readily soluble in concentrated muriatic acid. COMPOSITION OF METEORITES 155 ANALYSES OF RHABDITE P Fe Ni Co Cu Total Amount Fe+Ni-f- taken Co : P 1 16.35 4 8 - 8 S 33 ! 5 : 65 100.00 2.780: 2 iS-49 5i-io 32.99 0.42 100.00 0.2563 2.967: 3 15.46 56.71 27.36 0.47 .... 100.00 0.2772 2.984: 4 I S-3 2 55-3 28.78 0.60 .... 100.00 0.3276 3.015: 5 15.09 52.42 33.51 0.25 101.27 0.476 3.106: 6 15-05 41.54 42.61 0.80 .... 100.00 0.0986 3.053: 7 15.03 52.54 31.71 0.72 .... 100.00 0.5985 3.076: 8 14.86 46.22 37.98 0.96 .... loo.oo 0.2255 3-107:1 9 12.95 49-33 38.24 100.52 .... 3-672:1 REFERENCES 1. Santa Rosa. Coahuila: Wichelhaus, Pogg. Ann., 1863, cxviii, 633; glisten- ing needles insoluble in nitric acid. 2. Lime Creek. Cohen: A. N. H. Wien, 1894, ix, 115. Calculated to 100, after deduction of 1.54 per cent chromite and 2.62 per cent daubreelite; the analy- sis gave only 95.57 per cent. 3. Hex River Mountains. Cohen: A. N. H. Wien, 1894, ix, no. Calcu- lated to 100 after deduction of 0.53 per cent chromite and 0.68 per cent daubree- lite. 4. Sancha Estate, Coahuila. Cohen: A. N. H. Wien, 1894, ix, 106. Calcu- lated to 100, after deduction of 0.43 per cent chromite and 0.28 per cent carbon. 5. Bendego. Florence: Ann. Mus. Nac., Rio de Janeiro, 1896, ix, 182. Mixed with schreibersite. Trace of tin. Material not treated with salt of copper. 6. Beaconsfield. Sjostrom: Monatsber. Berlin Akad., 1897, 1041. Iron determined by difference. Cobalt determination incomplete. 7. Bolson de Mapimi. Cohen: A. N. H. Wien, 1894, ix, 103. Calculated to IOO, after deduction of 0.96 per cent residue and 2.15 per cent daubreelite. 8. Seelasgen. Cohen: Meteoreisenstudien, V, A. N. H. Wien, 1897, xii, 52. 9. Cranbourne. Flight: Ph. Tr., 1882, 891. Apparently quadratic, very brittle prisms, impervious to muriatic acid. Identified by Flight with rhabdite. COHENITE Lamprite in part Cohenite is a carbide of iron, nickel and cobalt, having the formula (Fe, Ni, Co) 3 C. It is found chiefly in the iron meteorites of the group of coarse octahedrites, having been identified in Beaconsfield, Bendego, Canyon Diablo, Magura, and Wichita County. It appears as silver-white, strongly magnetic and brittle crystals oxidizing to bronze yellow or tomback brown. Streak gray-black. Hardness 5.5-6. Sp. gr. 7.20-7.65, (Hussak 6.18). Cohenite is insoluble in dilute HC1 but is slowly dissolved by concentrated acid, and gives off in the latter process a petroleum-like odor. It is also soluble in copper ammonium chloride. 156 METEORITES It was first distinguished as a separate mineral by Wein- schenk in 1889, having been previously mistaken for schreib- ersite. It differs from schreibersite in being infusible and in giving no precipitate with ammonium molybdate. Crystals of cohenite are usually of elongated form and are often arranged parallel to the octahedral bands of their host. Definite forms have been described only by Hussak* who found them constituting crystalline aggregates in the Bendego meteorite. By dissolving the iron in weak acid these aggregates became separated and on the crystals so obtained Hussak identified the following isometric forms: 0(100), o(in), d(no), p(22i) and probably (311), (322) and (944). The habit of the crystals was tabular. In Ma- gura they reach a length of 8 mm. Carbides of composition similar or nearly similar to cohenite are found in artificial iron and in the native iron of Greenland. ANALYSES OF COHENITE Fe Ni Co C Total Fe+Ni+Co : C 1 94.34 0.13 5.53 100 3.67 2 9I-69 2.21 6.10 100 3-30 3 91.45 2.47 o.io 5.98 ioo 3.35 4 91.31 1.77 0.25 6.67 ioo 3.00 5 91.06 2.20 6.73 IOO 2.97 6 90 .94 2.22 0.30 6.54 IOO 3 . 06 7 90.80 2.37 0.16 6.67 ioo 3.00 8 89.81 3.08 0.69 6.42 ioo 3 . ii REFERENCES 1-2. Canyon Diablo. Florence: Am. Jour. Sci., 1895 (3), xlix, 105-106. I. Isolated crystals after deducting 3.64 per cent schreibersite. 2. Plates inter- grown with schreibersite. 3. Canyon Diablo. Tassin: Smithsonian Misc. Coll., 1907, 50, 212. Thin plates and rounded grains. Calculated to ioo after deducting .18 per cent schreibersite. 4. Canyon Diablo. Fahrenhorst: A. N. H. Wien, 1901, xvi, 375. Calculated to ioo after deducting 4.68 per cent of schreibersite. 5. Bendego. Dafert: Meteoritenkunde, Heft I, 117. Calculated to ioo after deducting 5.68 per cent schreibersite. 6. Beaconsfield. Sjostrom: Monatsber. Berlin Akad., 1897, 1043. Calculated to ioo after deducting 26.12 per cent schreibersite. 7. Wichita. Sjostrom: Zeitschriftfiir anorgan. Chemie, Hamburg and Leipzig, 1896, xiii, 57. Calculated to ioo after deducting 9.35 per cent schreibersite. 8. Magura. Weinschenk: A. N. H. Wien, 1889, iv, 95. Mean of three anal- yses calculated to ioo after deducting .65 per cent schreibersite. Traces of Cu-and Sn found. *Arch. Mus. Nac. Rio de Janerio, 1896, 9, 161-5. COMPOSITION OF METEORITES 157 MOISSANITE A silicide of carbon having the formula SiC was first found by Moissan* in the residue left after dissolving a 53 kg. piece of Canyon Diablo in hydrochloric acid, and treating this residue with hydrofluoric acid and boiling sulphuric acid. The mineral occurred in the form of small hexagonal crystals of a generally green color but varying from pale green to emerald green. Specific gravity 3.2. The mineral was unattacked by acids but gave potassium silicate on fusion with caustic potash and CO 2 on fusion with lead chromate. Kunzf suggested the name moissanite for the mineral in honor of its discoverer. In physical and chemical properties the mineral agrees with the previously known and artificially produced carborundum. Forty-four crystal forms have been identified on carborundum but none on moissanite. Moissan simply stated that the edges of the crystals observed by him were well-formed and the sides perpendicular. LAWRENCITE This name is applied to ferrous chloride, FeCl 2 , found sometimes in solid form but usually deliquescent in green drops on meteorites. The solid form has been reported only in the meteorites of Laurens, Smith Mountain, and Taze- well. These are all fine octahedrites with high percentages of nickel. Little description of the appearance of the sub- stance was given by the finders. It was simply stated that it occurred in crevices and became soft on exposure. The name lawrencite was given to the mineral by Daubree in honor of J. Lawrence Smith. J More frequent than the solid form of the mineral is the occurrence in the manner first described by Jackson in the meteorite of Limestone Creek. Jackson states that having washed the iron several times in distilled water he filed one side of it bright and left it exposed to the air. In a few days numerous drops of a *Comptes Rendus, 1904, 139, 778, and 1905, 140, 405-406. fAm. Jour. Sci., 1905, 19, 396-397. JC. R., 1877, 84, 69. Am. Jour. Sci., 1838, i, 34, 333. 158 METEORITES grass-green liquid collected on the surface of the iron and these soon became externally coated with a thin, brown film. The drops had a slightly alkaline, astringent taste but gave no alkaline reaction with turmeric paper. Qualitative tests showed the presence of iron, nickel, and chlorine. Quantitative analysis (reduced to percentages by Cohen*) gave: Fe Ni Cl 51.02 18.14 30.84 = 100 Fe+Ni : Cl = i.4 : i ~ Analysis of scraped material was also made by Jackson and another by Hayes but these analyses appear to have been much contaminated by foreign material. All show higher percentages of iron than normal ferrous chloride and thus indicate that impure material was used. The per- centages for pure ferrous chloride are: Fe Cl 44-i 55-9 On exposure to the air, lawrencite rapidly turns brown and becomes earthy, showing a change to ferric chloride (molysite), and ferric hydroxide (limonite). The reaction is: 6FeCl 2 +3O+3 H 2 O = Fe 2 O 6 H 6 + 4 Fe Cl, The ferric chloride is, however, reduced by contact with iron to form ferrous chloride again : 4 FeCl 3 +2Fe=6FeCl 2 so that the process is continuous. In addition there may occur a formation of free acid through hydrolysis of ferric chloride: 4 Fe Cl 3 +6 H 2 O = Fe 2 O 6 H 6 +6 HC1+2 Fe Cl, In connection with these chemical changes there is an increase of volume which causes splitting and disintegration of the meteorite. All lawrencite of meteorites shows qualitatively and quantitatively nickel present with the iron. This probably gives the green color to the mineral as compared with *Meteoritenkunde, 1894, Heft I, 231. COMPOSITION OF METEORITES 159 ferrous chloride which is colorless. The latter has been reported at Vesuvius. Meteoritic lawrencite is, then, a mixture of iron and nickel chlorides. It was early suggested that the lawrencite of meteorites was not a primary con- stituent, and might have been formed by the absorption of chlorine by the meteoric mass from the earth's atmosphere or the soil. This was the view of Shepard* and Mallet. | But as lawrencite varies in quantity in meteorites and ex- udes from pieces freshly cut from the interior of meteor- ites, there can be little doubt that it is an original con- stituent. It is known to be present, as has been mentioned, in clefts and hollows and it is also regarded by Cohen as distributed "in the intermolecular spaces. "J Its dis- tribution is indicated in a general way in a meteorite by areas inclined to rust. The borders of accessory con- stituents are, as is well known, especially liable to such a change. Cohen thinks it probable that the lawrencite of a meteorite gradually works outward by diffusion. He bases this view upon the fact that in the unaltered interior of Forsyth he found 0.17 per cent of chlorine but in the rust crust 4.99 per cent. Nevertheless it is highly probable also that the lawrencite is more or less irregularly distributed. In Deep Springs the non-rusted portion showed 0.016 per cent chlorine, and the easily rusting portion 0.99 per cent, while Venable found 0.39 per cent in a mass analysis. In Deep Springs, as in the Cape iron and Lick Creek, the easily rusting portions are divided by rather sharp boundaries from those which do not rust. Probably through a process of diffusion to the surface and thus of escape, the lawrencite of a meteorite often seems to disappear after a time. This exhaustion is shown by a cessation of the tendency to rust. Cohen describes such a cessation in a section of the Cape iron after a period of 15 years, and in one of Sao Juliao after a much shorter time. The presence of chlorine, indicating lawrencite, is often reported in iron meteorites in appreciable percentages by *Am. Jour. Sci., 1842, i, 43, 359-362. fAm. Jour. Sci., 1871, 3, 2, 14. |Meteoritenkunde, 1903, Heft II, 266. 160 METEORITES analysts, the largest per cent found being 1.48 per cent, reported by Jackson in Limestone Creek. According to experiments made by Cohen,* the extraction of chlorine for analysis can best be made by digesting a few grams of the meteorite in dilute sulphuric acid. Dilute nitric acid or boiling water is less effectual for this purpose. It is to the presence of lawrencite in meteorites that their fre- quently observed "sweating" is doubtless to be ascribed. Over the surface of such meteorites the continuous forma- tion of liquid drops may be observed and relatively rapid decomposition of the mass takes place, at least until a protective crust is formed. Conspicuous examples of such meteorites are Cranbourne and Toluca. Other meteorites, such as St. Genevieve, some individuals of Canyon Diablo, etc., show no tendency to sweating or rusting although in a fall consisting of numerous individuals, such as Toluca and Canyon Diablo, there is marked difference in the rusting tendency of different individuals. Lawrencite seems to be almost wholly associated with nickel-iron and to be con- fined, therefore, largely to the iron meteorites. "Sweating" of some iron-stone and stone meteorites indicating the prob- able presence of lawrencite in them has, however, been ob- served. Such meteorites include Crab Orchard, Hainholz, Morristown, and Sierra deChaco among the iron-stones and Charwallas, Marion, and Nagaya among the stones. Other salts besides lawrencite soluble in water have been found in some meteorites, chiefly carbonaceous ones, the porous texture of which suggests absorption or formation of these salts from the earth's atmosphere rather than their existence as primary constituents. The soluble salts so found include chlorides and sulphates of sodium, potassium, ammonium, magnesium, and calcium. They are usually found in the water extract of the meteorite but occasionally are obtained as sublimates from heating the meteorite in powdered form. WATER Water is as a rule conspicuous by its absence from mete- orites, yet in some occurs in appreciable quantities. As *Meteoritenkunde, 1903, Heft II, 266 COMPOSITION OF METEORITES 161 the meteorites in which it is chiefly found are of a porous nature, some authorities are inclined to regard its origin in meteorites as always terrestrial. It is perhaps never a primary constituent and yet some meteorites which have been picked up immediately after their fall show a rusting of the interior which would seem pre-terrestrial. In the car- bonaceous meteorites Alais, Cold Bokkeveld, Nagaya, and Orgueil, water has been found in quantities of from 6 to II per cent. From these it is obtained by heating to a tem- perature above 100 C. In other stony meteorites such as Bishopville,L'Aigle, and Pultusk, water has been obtained in quantities of from o.i to 1.43 percent by heating the stone to redness. Part of such water doubtless exists in com- bination, as for example, with iron forming iron hydroxide and it has been suggested that part may be formed from hydrocarbons by heating. No mineral in meteorites except limonite (if that be an original constituent) possesses water. Pisani observed that the powder of the Orgueil meteorite which lost 9.15 per cent water by drying, took up 7 per cent again in a few hours. QUARTZ On the whole, quartz is conspicuous by its absence from meteorites. It has, however, been identified in several irons, though almost wholly in their superficial portions. This has caused some authorities to doubt whether quartz is ever an original constituent, but there are reasons for believing that it is so. The first satisfactory observation of quartz in meteorites was by Rose* of grains which he observed in the crust of one of the Toluca irons. These were further studied in 1895 by Laspeyresf who found in the earthy, loamy crust of a Toluca individual weighing 10 kilos numerous brilliant quartz crystals reaching a _size of 2 mm. The crystals showed the forms oo R (1010), R (ion) and R (oili). They also Rad the usual characters of siliceous meteoritic minerals in being fissured and brittle and in showing rounded *Monatsber. Berlin Akad., 1861, 406-407. fZeitschrift fur Kryst., 1894, xxiv, 485-488. 162 METEORITES edges and solid angles. The specific gravity was 2.65. Index of refraction about that of Canada balsam. High interference colors. Silica skeleton with salt of phosphorus. Soluble in hydrofluoric acid. Unaltered by strong heating in the oxidizing flame. Other occurrences of quartz have been noted in the insoluble residues of meteoric irons. Thus Joy found in the insoluble residue of Cosby Creek some white, opaque, and some transparent grains which would scratch glass but not quartz and which he, therefore, re- garded as quartz. Similar residues were found by Cohen and Weinschenk in all the irons they investigated. These included Beaconsfield, Bischtiibe, Glorieta Mountain, Hex River Mountains, Ivanpah, Kokstad, Lime Creek, Locust, Magura, Misteca, Rasgata, Schwetz, and Toluca. The grains found were colorless and transparent and averaged less than o.i mm. in diameter. From the pyrrhotite of Caille and Charcas apparent quartz grains were isolated by Daubree and from the schreibersite of Sao Juliao by Cohen. In none of these occurrences has the quartz been found in place, hence its existence as a fundamental consti- tuent of meteorites is not altogether certain, but indications point strongly to a pre-terrestrial origin for the grains found. TRIDYMITE Asmanite Tridymite has been identified in the iron-stone meteorite of Steinbach and its probable presence has been reported in Vaca Muerta'and Crab Orchard, which are also iron-stone meteorites, and in the stone meteorite Fisher. In Steinbach the tridymite occurs as colorless, rounded grains or plates which reach 3 mm. in their largest dimension. They are generally stained with iron superficially and like other meteoritic minerals are much rounded. The grains also are brittle and have a resinous luster. The hardness is given as 5-5 by Maskelyne, 6.5 by Rath. Specific gravity 2.24- 2.27. Maskelyne regarded the mineral as orthorhombic and considered it, therefore, a new species to which he gave the name asmanite from dsman, the Sanskrit word for COMPOSITION OF METEORITES 163 thunderbolt.* Assuming an orthorhombic form for the mineral Maskelyne distinguished twelve forms, the equiva- lents of which in the now generally accepted hexagonal system have not as yet been determined. The existence of tridymite in a meteorite shows, according to the researches of Wright and Larsen,t that it has been heated to a temperature above 800 C, since, tridymite forms between 800 and 1625 C. MAGNETITE Several stone or iron-stone meteorites have been found to contain black, magnetic grains which dissolved in hydro- chloric acid without effervescence to form a yellow solution. In the meteorites of Shergotty and Dona Inez these are sufficiently abundant to form an essential constituent, in Shergotty constituting 4.57 per cent. Similar grains occur as inclusions in maskelynite, pyroxene, and chrysolite in the above and other meteorites. They are regarded as magnetite. No well-marked crystals of meteoritic magnetite have as yet been described. In several iron meteorites magnetite has been reported as a constituent but few analyses have been made. Tassinf observed a chromiferous magnetite occurring as rounded grains of blue-black color and dull luster in Canyon Diablo, associated with "troilite and silicon compounds in areas rich in carbon." Analysis gave: Fe 2 O 3 65.25 FeO 30.05 Cr 2 O 3 5 . 20 100.50 Meunier analyzed magnetite from the crust of Toluca obtaining the material by first treating the crust with *Phil. Trans., 1871, 161, 361. fAm. Jour. Sci., 1909, 4, 27, 423. JProc. U. S. Nat. Mus., 1908, 34, 687. 164 METEORITES chloride of mercury, then with very dilute HC1 and then with magnets. His analysis gave: Fe 2 O 3 68.93 FeO 28.12 NiO 2 . oo CoO tr. 99.05' That the crust formed upon the surface of iron meteorites in passing through the air has the composition of magnetite was also shown by Farrington* from an analysis of such a crust upon the Quinn Canyon meteorite by Nichols. The analysis calculated to 100 after deduction of extraneous compounds gave: Fe 2 O 3 72-38 FeO 27.62 100. CHROMITE Chromite is a common constituent of meteorites, being not infrequent in the irons and almost universally present in the stones. It occurs in the stones usually in the form of grains, often of microscopic dimensions, though sometimes as large as a pea. In thin sections it can usually be readily recognized by its translucent ruby to purple color and iso- tropic characters. In the irons it usually occurs in nodules, sometimes of considerable size, as in the case of one of the Coahuila irons, where one of oval form found by Smith measured 5x7 inches (12 x 17 mm.) in diameter. This was a black, granular mass, feebly translucent and of dark red- dish-purple color. In other irons, as in Carthage, Schwetz, and Bendego, chromite occurs closely associated with pyr- rhotite and cohenite. It is usually in octahedral crystals averaging about one millimeter in diameter. From Bendego highly modified crystals were obtained by Hussak,f and the following forms determined: in, no, 100, 553, 774, 221, *Pubs. Field Museum, 1910, Geol. Ser., iii, 176. fAnn. Mus. Nac. Rio de Jan., 1896, 9, 165-171. COMPOSITION OF METEORITES 165 552, 331, 441, 211, 311, 210, 310, 510. Only a few of these forms occur in terrestrial chromites. The octahedron pre- dominated, but the crystals were often tabular from the large development of two of the planes. In Ensisheim Shepard identified the forms in and no; in Lodran Lang observed in, no, and 311; and in Greenbrier County Fletcher reported in, no, and 221. Borgstrom* found well-defined crystals about 2 mm. in diameter at the boun- dary between the chrysolite and nickel-iron of the Marja- lahti meteorite. The dominant forms were in and no. Combined with these were 331, 311, and 551 of which the latter had not previously been noted on the mineral. The physical and chemical properties of meteoritic chromite do not differ essentially from those of terrestrial chromite ex- cept that meteoritic chromite is sometimes probably decom- posed by acids. This is assumed from the content of Cr 2 O 3 often found in the soluble portion of meteorites. The color of meteoritic chromite is in general black, with sub-metal- lic luster; streak brown; non-magnetic or only weakly so; sometimes very brittle. Analyses of meteoritic chromites show the presence of considerable alumina and magnesia. ANALYSES OF METEORITIC CHROMITE Cr 2 O 3 A1 2 O 3 Fe 2 O 3 FeO MgO Total 1 65.63 3.78 .... 25.84 4.27 NiO 0.73 100.25 2 65.49 33-00 0.40 Si O 2 0.50 99-39 3 65.01 9.95 18.97 5.06 98.99 4 64.91 9.85 17.97 4-96 SiOai.38 99.07 63.40 5.30 26.30 5.00 100. 6 62.71 33.83 .... 96.54 7 62.00 .... 41.00 .... .... 103. 61.39 1-96 30.46 6.70 100.51 9 59-85 .... 27.93 12.22 100. 10 56.82 11.36 26.14 5-68 100. 11 56.70 12.38 27.60 4.00 100.68 12 52.13 10.25 37-68 100.06 13 39-40 28.50 31-50 0.60 loo. 14 24.60 54-50 20.90 100. REFERENCES 1. Marjalahti. Borgstrom: Geol. Foren. i. Stockholm Forh., 1908, 30, 331. Crystals, 2 mm. in diameter. 2. Admire. Tassin: Proc. U. S. Nat. Mus., 1908, 34, 686. Non-magnetic, jet-black grains of brilliant luster. *Geol. Foren. i. Stockholm, Forh., 1908, 30, 331. 166 METEORITES 3. Mount Vernon. Tassin: Proc. U. S. Nat. Mus., 1908, 34, 685. Brilliant black crystals. 4. Mount Vernon. Tassin: Proc. U. S. Nat. Mus., 1908, 34, 686. Minute, rounded, brownish-black grains. 5. Canyon Diablo. Tassin: Proc. U. S. Nat. Mus., 1908, 34, 688. Small, black, octahedral crystals and rounded grains. 6. Coahuila. Smith: Am. Jour. Sci., 1881, 3, 21, 462. Large nodule. 7. Bjurbole. Ramsay and Borgstrom: Bull. Comm. Geol. de Finlande, 1902, 12, 13. Black powder. 8. Marjalahti. Tassin: Proc. U. S. Nat. Mus., 1908, 34, 687. Brilliant, blue- black crystals. 9. Klein-Wenden. Rammelsberg: Ber. Berlin Akad., 1844, 245. 10. Shalka. Foullon: A. N. H. Wien, 1888, 3, 199. 11. Allegan. Tassin: Proc. U. S. Nat. Mus., 1908, 34, 688. Blackish-brown grains. 12. L'Aigle. Schwager: Sitzber. Miinchen Akad., 1878, 8, 39-40. 13. Sewrukof. Eberhard: Arch. f. d. Nat. Lib. Ehst. Kurl., 1882, 9, 137. 14. Lodran. Rammelsberg: Abh. Berlin Akad., 1870, 93. Cohen states that this analysis is of doubtful accuracy Meteoritenkunde, i, 246. BREUNNERITE Determination of carbonates in meteorites has thus far been confined to the observation of small, transparent crys- tals which were isolated from the meteorite of Orgueil. These had characters which indicated that they were breunnerite, iron-magnesium carbonate (Mg, Fe) CO 3 . The crystals were of rhombohedral form, with angles of IO5-IO7 and exhibited cleavage in three directions. They showed weak, pearly luster. They dissolved slowly in cold HC1 and showed qualitatively iron, magnesia and carbonic acid.* The crystals were of small size (]4~Y\ mm.) and few in number. As Orgueil is a very porous meteorite it has been suggested that this carbonate may have been formed by the action of the terrestrial atmosphere, but as some crystals were found in the interior of the meteorite it is possible that they were of primary origin. FELDSPAR Minerals of the feldspar group are common constituents of the stone meteorites, though less abundant than chrysolite and enstatite. They are chiefly prominent in the groups of eukrites and howardites, and among the silicates of the grahamites. Small percentages of alkalies found in the chondrites are usually regarded as indicating the presence of *Pisani: Compt. Rendus, 1864, 59, 135. COMPOSITION OF METEORITES 167 feldspars in them though the feldspars are difficult to detect by optical means. The most common and best defined feldspar occurring in meteorites is anorthite. As seen with the naked eye, especially in the eukrites, it is usually nearly opaque, dull, and of a snow-white color. In some meteorites, however, it becomes more or less transparent and has a marked vitreous luster. It occurs in the form of crystals, grains, and splinters and these are usually of appreciable size. One crystal in Jonzac measured I cm. in length. The crystals are usually more or less lath-shaped. The FIG. 51. Forms of anorthite from the Juvinas meteorite. After Rose, Lang and Tschermak. meteorites in which the feldspars are abundant usually exhibit the ophitic structure which is frequently observed in diabase. As in terrestrial rocks, this structure is due to the fact that the feldspars crystallized before the pyroxenes. Terminated crystals of anorthite have been found in druses of Juvinas and have been measured and described by Rose, Lang and Tschermak. Their forms are shown in the ac- companying figure (Fig. 51) of which the first shows a twin, the second a tabular habit, and the third a Carlsbad twin. The forms observed were: M oio P ooi e 02 1 T' TTo 1' Tio X 101 O III p' III Meteoritic anorthite is soluble in hot HC1. It has been isolated and analyzed only in the eukrites and grahamites. The following analyses have been made: 168 METEORITES ANALYSES OF ANORTHITE Si0 2 A1 2 O 3 Fe 2 O 3 CaO MgO Na 2 O K 2 O Total 44-38 46.19 33.73 31.26 3-29 2-93 18.07 16.98 0.36 I . 12 1.03 I.I4 0-33 0.50 101 . 19 IOO. 12 42.91 36.76 17-56 2-77 IOO.OO 42.87 36.59 17-50 2.04 i .00 100.00 42.02 37-77 16.41 0.96 97.16 REFERENCES 1. Juvinas. Rammelsberg: Pogg. Ann., 1848, 73, 588. Isolated crystals. 2. Stannern. Rammelsberg: Pogg. Ann., 1851, 83, 592. Isolated crystals. 3. Petersburg. Calculated by Rammelsberg from the alumina and alkali of Smith's mass analysis. Abh. Berlin Akad., 1870, 129. 4. Frankfort. Calculated by Rammelsberg from the alumina and alkali of Brush's analysis. Abh. Berlin Akad., 1870, 131. 5. Morristown. Merrill: Am. Jour. Sci., 1896, 4, 2, 151. Analysis of 0.19 gram separated by heavy solution. Na 2 O not determined. Under the microscope, meteoritic anorthite usually shows a large extinction angle, 3O-38. Microscopic inclusions are common. These may be either of rounded or elongated form and may be colorless to brownish. Many are doubt- less glass inclusions. They usually show some regularity of arrangement. Anorthite is estimated to constitute about 35 per cent of the meteorites of Juvinas and Stannern and from 20 to 30 per cent of those of Petersburg and Frankfort. In the residue from the solution in nitric acid of Bischtiibe, a meteoric iron, Kislakowski* found about 8 per cent of siliceous grains which he regarded as anorthite. Chondri made up wholly of anorthite have been observed in several meteorites. Feldspars other than anorthite have rarely been isolated from metorites and their presence can usually be more readily inferred than proven. Where seen, their appear- ance under the microscope is to be distinguished from that of anorthite by their narrower and more abundant twinning lamellae and smaller extinction angle. They are not attacked by hot, concentrated HC1, and hence are found on analysis in the portion insoluble in acids. Like anorthite they occur as laths, grains, and splinters. As laths they are especially well seen in some of the grahamites, having been observed in Vaca Muerta and Crab Orchard. Where feldspar is not very abundant as a constituent it commonly * Bull. Soc. Imp. de Moscou, 1890, No. 2, 190-197. COMPOSITION OF METEORITES 169 occurs in grains, often of small size and not even showing twinning lamellae. In such cases the grains exhibit un- dulatory extinction. Plagioclase grains were isolated by Prendel from Zabrodje and Grossliebenthal which showed in Zabrodje extinction angles of 12 and 2, indicating felds- par of the composition Ab 6 Ani and in Grossliebenthal angles of 8 and i indicating feldspar of the composition, Ab 4 Ani. In the residue after dissolving the Toluca iron in HC1, Laspeyres found plagioclase grains showing twinning lamel- lae. Plagioclase other than anorthite has not been isolated from meteorites for chemical analysis but the following com- positions have been calculated from mass analyses: ANALYSES OF METEORITIC PLAGIOCLASES Si0 2 A1 2 3 CaO Na 2 K 2 O Total i 64.97 22.06 3.01 9.96 100 2 63.50 22.20 4.00 9-20 I. 10 100 3 61.85 24.09 5.25 8.81 ioo 4 53-17 29.51 11.55 4-33 i-44 ioo REFERENCES 1. Hessle. Lindstrom: Bihang Svenska Vet. Akad. Stockholm, 1869, 8, 723. 2. Hvittis. Borgstrom: Die Met. von Hvittis u. Marjalahti. Helsingfors., 1903, 32. Calculated to ioo after deducting intermixed chrysolite. 3. Gopalpur. Tschermak: Sitz. Wien Akad., 1872, 65, I, 143-144. 4. Tennasilm. Schilling: Archiv. fur Naturkunde Liv. Ehst. u. Kurlands, 1882, i, 9, 113. The first three feldspars are oligoclase, having the re- spective compositions, Ab 6 An, Ab 4 An, and Ab 3 An. No. 4 is labradorite, Ab 2 An 3 . The feldspars of a large number of meteorites were studied optically by H. Michel.* He found the feldspar of the eukrites and howardites to be largely anorthite while that of the chondrites was chiefly oligoclase. Considering the chondrites alone, the white chondrites were found to have abundant feldspar while the black chondrites were free from it. Other chondrites had varying amounts according as they approached either one of these classes. In the chondrite of Waconda containing light and dark por- tions the amount of feldspar differed in each as if they were separate meteorites. In many chondrites in which the pre- sence of a feldspar was indicated by chemical analysis, glass took the place of the expected feldspar. It was apparent, *Tsch. Min. u. Pet. Mitt., 1912, 31, 563-658. 170 METEORITES therefore, that the glass had the composition of feldspar of which it may have been, according to this author, an alter- ation product, although a primary origin is entirely possible. MASKELYNITE A mineral of hitherto unknown characters was found by Tschermak* making up 22^/2 per cent of the meteorite of Shergotty. It was colorless, isotropic, transparent, of vitreous luster, and conchoidal fracture. Hardness 6; specific gravity 2.65. Before the blowpipe it fused in thin splinters to a clear, colorless bead. The fine powder was slightly decomposed by hydrochloric acid. The mineral occurred in feldsparlike laths bounded by straight lines. Lines parallel to their lengths gave these laths an appearance like plagioclase, but the cleavage of plagioclase was lacking. Analysis of the mineral after deduction of 4.7 per cent iron oxide which was ascribed to included magnetite, gave, when calculated to 100, the composition: SiO 2 A1 2 O 3 CaO Na 2 K 2 O 56.3 25.7 n.6 5.1 1.3 = 100 This corresponds to a plagioclase composed of. 51.75 per cent albite and 48.25 per cent anorthite, if K 2 O be omitted and the analysis be calculated to 100. To this mineral Tschermak gave the name maskelynite in honor of N. Story-Maskelyne. Tschermak regarded it a fused feldspar At the same time by Foullon and later by several observers a similar mineral was found forming an accessory constituent in many of the chondrites. It was described by Cohen, f in Madrid, as being in the form of small, elongated or rounded particles lying among the essential minerals and. taking part in the constitution of the chondri. The grains had a diameter of o.i mm. or more, were irregularly bounded, had weak double refraction and occasional indulatory extinction. The index of refraction was about that of Canada balsam. *Sitz Wien Akad., 1872,65, 127-131. fMitth. nat. Verein Neu-Vorp. u. R., 1896, 28, 103-105. COMPOSITION OF METEORITES 171 Winchell* found in Fisher in the midst of an isotropic substance, portions which were doubly and more highly refracting and which showed lamellae resembling those of plagioclase. The axial angle was 15. Both the isotropic and doubly refracting portions gave the same qualitative composition. Winchell regarded the isotropic substance as glass from which the anisotropic mineral had crystallized. This anisotropic mineral he regarded as maskelynite. While Tschermak regarded maskelynite as a fused feldspar, Groth and Brezina consider it a distinct mineral allied to leucite. Its exact nature is yet to be determined. ORTHORHOMBIC PYROXENES ENSTATITE, BRONZITE AND HYPERSTHENE Chladnitt, Piddingtonite, Fictorite, Shepardite Orthorhombic pyroxenes rank third in quantity among the constituents of meteorites, being exceeded only by nickel- iron and chrysolite. Together with chrysolite they form almost the entire substance of the great group of chondritic stone meteorites and are also an important ingredient in the howardites, rodites, amphoterites, mesosiderites, and gra- hamites. Enstatite constitutes one meteorite, Bishopville, almost alone, and hypersthene is practically the sole con- stituent of Manegaum, Ibbenbiihren, and Shalka. These pyroxenes are colorless to snow-white in enstatite and pre- sent various shades of gray, green, and brown in bronzite and hypersthene. Inclusions may also darken the color. Often a color like that of chrysolite is exhibited and partly for this reason the recognition of orthorhombic pyroxene was not made until late in the study of meteorites. Other distinguishing characters of the orthorhombic pyroxenes as seen in meteorites are a frequent fibrous structure, prismatic cleavage, pinacoidal parting, straight extinction, low inter- ference colors, prismatic habit, and little or no solubility in acids. The first to detect the occurrence of orthorhombic pyroxene in meteorites was Langf who determined the * Amer. Geol., 1897, 20, 316-317. fSitz. Wien Akad., 1869, 59, 2, 848-856. 172 METEORITES pyroxene in Steinbach previously regarded as monoclinic to be orthorhombic. By his investigations and those of others the following crystal forms have been determined on the orthorhombic pyroxenes of meteorites: a..ioo 77.. 140 g. .021 ^..252 b . .010 ..320 fl. .031 C..2I2 C..OOI z..2io ^..201 y. .432 m..no 8. . 520 /. .502 i..2ii a. .230 X..3IO O..III ^..623 X . .221 f.. 4 I2 FIG. 52. Enstatite from the ' Steinbach meteorite. n..l2O and reddens litmus paper. Decomp. by HC1 and HNO 3 . Easily soluble in HC1 with evolution of H 2 S. Sol. in HNOs with evolution NO-> Strongly magnetic before heating: Gives no odor on roasting nor reddening of litmus paper. Heated with magnesium wire in a closed tube and moistened with water gives the disagreeable odor of phosphuretted hydrogen. Soluble in HC1 and Aqua regia. Insol. in copper ammonium chloride. Treated with HNOs or dis- solved in HC1 gives yellow ppt. with Am MoO 4 . Decomp. with difficulty bv HNO 3 Schreibersite (Fe, Ni, Co GENERAL CHARACTERS SPECIFIC CHARACTERS Reacts for Malleable nickel Sol. in HC1 or HNO 3 Nickel-iron Fe+\i Kamacire Ke u Usually in flexible plates which fuse on Taenite thin edges B. B. Sol. in copper ammonium chloride Fe, N.+CY Fe Ni+( Brittle. Fuses in R. F. emitting sparks. Cohenite Decomposed with difficulty by con- cen. HC1. Sol. in copper ammonium chloride (Fe, Ni, Cc Fine powder slowly but completely soluble in HC1 Magnetite. FegO Not magnetic before heating Impart a green color to salt of phosphorus or borax bead (Chromium) Reacts for sulphur when roasted in open tube. Becomes magnetic \\lun heated in R. F. Brittle. Sol. in HNOs and Aqua Regia Daubreelite A little of the fine powder when mixed with an equal volume of Na 2 CO.j and intensely heated on charcoal gives a magnetic mass. Insol. in acid Chromite Soft. Soils the fingers. Very refractorv. Graphite Deflagrates with KNO 3 . Use equal parts mineral and KNO :i . RALS WITH METALLIC LUSTRE ;IBLK Color Sneak . Cleavage and Fracture Hard- ness Spec. Grav. Fusi- bility Crystalli- zation Bronze- ( iravish- Uneven 4 4.68- 2-5-3 Hexagonal yellow to black 4.82 brown Tin-white, steel-^rav. Grayish- black Brittle 6 - 5 6.3- 7.28 2-5 bronze- yellow Steel-gray Steel-gray Octahedral Cubic Hackly 4-5 7-3- 7-5 Isometric Steel-gray 7 .8c^ 7.88 Isometric Tin-white to golden- yellow Isometric Tin-white, bronze- Yell ovv 5.5-6 7-23- 7.24 Isometric Iron-hlack Black Parting octahedral. Uneven 6 5.18 Isometric Black Black One direction 5-01 Massive Scaly Iron-hlack to brownish black . Dark brown Uneven 5-5 4.6 Isometric U. massive Iron-black Black Basal (perfect) 1-15 2.20 Hexagonal Rh. Foliated GENERAL INDEX Achondrites, 108, 198 Aerolites, 197 Aerolitics, 5 Aerosiderites, 197 Aerosiderolites, 197 Allen, E. T., 141, 176 Aluminum, 113 American Museum collection, 225 Amherst collection, 223 Amorphous carbon, 124 Amphoterite, 198 . Analyses of anorthite, 168 of chromite, 165 of chrysolite, 186 of cohenite, 156 of gases, 191, 192, 194, 195, 196 of hedenbergite, 179 of insoluble silicates, 175 of kamacite, 132 of orthorhombic pyroxene, 173 of plagioclase, 169 of pyrrhotite, 145 of rhabdite, 155 of schreibersite, 153 of taenite, 134 of vein material, 87 Angrite, 198 Anorthite, 167 Ansdell, G., 124, 194 Antimony, 116 Apatite, 187 Apparent paths of meteors, 32 Argon, 113 Armored surfaces, 86 Arsenic, 116 Asiderites, 197 Asmanite, 162 Ataxites, 98, 203 Atmospheric resistance, 18 Augite, 105, 1 80 Average composition of meteorites, 216 Balkeneisen, 130 Ball, Robert, 33 Bandeisen, 133 Barium, 116 Barnard, E. E., 212 Barringer, D. M., 25 Bell-shaped meteorites, 70 Bent figures, 91 Berlin collection, 223 Berwerth, F., 5, 35, 96, 98, 99, 178, 181, 184, 187, 222 Berzelius, J. J., 124 Bismuth, 116 Blake, W. P., 74 Borgstrom, L. H., 21, 83, 139, 150, 165 Breunnerite, 166 Brezina, A., 77, 83, 89, 101, 108, no, in, 121, 122, 126, 130, 131, 139, 141, 143, 149, 150, 171, 198, 222, 223 Brezina lamellae, 100, 150 British Museum collection, 222 Bronzite, 171 Bustite, 198 Calcium, 113 Carbon, 113, 124 Carbon dioxide, 191, 196 monoxide, 191, 196 Chamberlin, R. T., 195 T. C., 211, 212 Chassignite, 198 Chondri 102, 127 Chondrites, 199 Chondritic structure, 102, 206 Chladnite, 171, 198 Chlorine, 113, 159 Chromite, 164 Chromium, 113 Chrysolite, 103, 181 Classification, 197 Cliftonite, 122 Clinoenstatite, 176 Clinohypersthene, 176 Cobalt, 114 Codonoid shape, 70 Cohen, E., 50, 68, 79, 88, 89, 90, 120, 123, 126, 134, 140, 147, 149, 151, 152, 159, 160, 162, 172, 174, 175, 177, 178, 179, 1 86, 1 88 Cohenite, 155 Column shape, 74 Combs, 94 Comets, 208 Cone-shape, 60 Coon Butte, 23 Copper, 114 Cricoid shape, 74 227 228 GENERAL INDEX Crust of iron meteorites, 78 of stone meteorites, 82 zones, 82 Cubic meteorites, 97 Daily falls, 40 Darwin, G. H., 211 Daubree, A., 51, 52, 105, 152, 157, 162, 187, 197 Daubreelite, 144, 146 Davison, J. M., 136 Derby, O. A., 120, 152 Dewar, J., 124, 194 Diallage, 177 Diamond, 118 Diopside, 178 Direct motion, 43 Disruptive forces, 212 Distribution of showers, 50 Dyslytite, 147 Eastern hemisphere, meteorites of, 35 Edmondsonite, 133 Eisenglas, 79 Elements, 113 Enstatite, 104, 171 Etching figures, 127 Eukrites, 109, 198 Eutectic, 137 Faulting, 90 Feldspar, 166 Ferrous chloride, 157 Field Museum, 223 Fletcher, L., 122, 126, 137, 165 Flight, W., 193, 223 Foote, A. E., 119 Foote Mineral Co., 128 Forms of meteorites, 60 Forsterite, 181 Front side of meteorites, 63 Fiilleisen, 136 Gases, 190 General characters, I Geographical distribution, 34 Gilbert, G. K., 25 Glanzeisen, 147 Glass 105, 190 Gnathoid shape, 75 Gold, 116 Graham, T., 191 Grahamite, 201 Graphite, 123, 144, 194 Haidinger, W., 45, 64, 74, 121, 140 Harnischflache, 86 Harvard collection, 223 Hedenbergite, 179 Helium, 114 Herschel, 21 Hexahedrites, 98, 203 Hobbs, W. H., 68 Hourly falls, 41 Howardites, 35, 199 Huntington, O. W., 97, 119, 122 Hussak, E., 156, 164 Hydrocarbons, 188 Hydrogen, 114, 191 Hypersthene, 171 Injuries. caused by meteorites, 28, 29 Iridium, 114 Iron, meteoric, 114 terrestrial, 206 Jaw-shaped meteorites, 75 Kamacite, 93, 129, 130 Kunz, Geo. F., 65, no, 118, 119, 149, 157, 183, 225 Labradorite, 169 Lamprite, 147, 155 Lane, A. C., 177 Lang, V. v., 167, 171, 172, 174 Lawrencite, 157 Lead, 116 Linck, G.,97 Lithosiderites, 201 Liversidge, A., 70, 72, 116 Lockyer, J. N., 210 Lodranite, 201 Lowell, P., 44 Magnesium, 114 Magnetite, 163 Manganese, 114 Mallet, J. W., 191 Maskelyne, N. S., 5, 138, 139, 162, 178, ivyr , 19 , 7 '" 3 Maskelynite, 105, 170 Merrill, G. P., 139, 188 Mesosiderites, no, 201 Meteorin, 133 Meteorite collections, 220 Meteor Crater, 23, 25 Meteorite showers, 46, 50 Meteoritic dust, 219 Meteoritic hypothesis, 210 Meteoritics, 5 Methods of etching, 127 Meunier, S., 88, 140, 147, 151, 163, 197, 223 Michel, H., 169 Minerals of meteorites, 116 GENERAL INDEX 229 Moissan, H., 120, 121, 157 Moissanite, 157 Monoclinic pyroxenes, 177 Monthly falls, 38 Monticellite, 188 Moon, 25, 207 Moulton, F. R., 32 Naming of meteorites, 4 Neumann, J. G., 96, 97 Neumann lines, 96 Newton, H. A., 44, m, 208, 213 Ngawite, 200 Nickel, 3, 114 Nickel-iron, 105, 124 Niessl, v., 19 Nitrogen, 115, 191 Number of individuals in showers, 49 of falls, 4 Observation of metec rs, 29 Octahedral figures, 92 Octahedrites, 202 Oldhamite, 138 Oligoclase, 169 Ophitic structure, 109, no, 167 Orbits of meteorites, 213 Origin of chondri, 213 of meteorites, 205 of showers, 50, 52 of veins, 88 Ornansite, 200 Orthorhombic pyroxenes, 171 Orvinite, 200 Osbornite, 139 Ostracoid shape, 69 Oxygen, 115 Palladium, 115 Pallasites, 201, 202 Paris collection, 223 Partsch, P., 121, 126, 222 Partschite, 147 Patton, H. B., 177 Peltoid shape, 69 Phenomena of falling meteorites, 7 Phosphornickeleisen, 147 Phosphorus, 115 Physical effects, 217 Pickering, W. H., 40, 44, 213 Piddingtonite, 171 Fittings, 63 Plagioclase, 105, 169 Platinum, 115 Plessite, 93, 130, 136 Potassium, 115 Pyroxene, 171 Pyrrhotite, 83, 140 Quantitative classification, 204 Quartz, 161 Radium, 115 Ramsay, W., 83 Rath, G. v., 64, in Rear side of meteorites, 63 Reichenbach, C. v., 79, 88, 89, 93, no, D ;"' I2 g^3o, 131, 136, H3 Reichenbach lamellae, 100, 143 Relation of meteorites to comets, 208, 213 Retrograde motion, 43 Rhabdite, 147 Ring shape, 74 Rinne, F., 95, 138 Rodite, 198 Rose, G., in, 121, 122, 140, 142, 161, 167, 180, 182, 183, 198, 223 Ruthenium, 115 Saturn, 211 Schiaparelli, J. V., 19, 213 Schreibersite, 144, 147 Secondary crust, 84 Shell shape, 69 Shepard, C. U., 5, 165, 188, 223 Shepardite, 171 Shergottite, 109, 199 Shield shape, 68 Showers, 46 Siderites, 197 Siderolites, 201 Siderophyr, 201 Siemaschko, J. v., 126 Silicon, 115 Size of meteorites, 54 Slickensided surfaces, 89 Smith, J. Lawrence, 123, 142, 146, 152, 164, 189, 223 Smith, S. W. J., 133 Sodium, 115 Soluble salts, 160 Sound phenomena, 20 Sporadosiderites, 197 Strontium, 116 Structure of iron-stone meteorites, 101 of meteorites, 92 of stone meteorites, 102 Styloid shape, 74 Sulphur, 116 Syssiderites, 197 Tadjerite, 200 Taenite, 93, 129, 133 Tassin, W., 139, 163 Terrestrial relations, 214 Tesseral octahedrite, 95 230 GENERAL INDEX Times of fall, 37 Tin, 116 Titanium, 116 Travers, M. W., 196 Tridymite, 162 Troilite, 140 Tschermak, G., 64, 82, 83, 86, 88, 104, 105, 106, 107, 108, 109, 112, 143, 149, 167, 170, 171, 179, 180, 184, 185, 188, 198 United States National Museum Col- lection, 225 Vanadium, 116 Veins, 85 Velocity of meteorites, 19, 22, 44 Victorite, 171 Vienna collection, 222 Wadsworth, M. E., 108 Wahl, W, 176, 177 Ward, H. A., 55, 225 Warmth of meteorites, 20, 27 Water, 160 Weinbergerite, 181 Weinschenk, E., 120, 123, 149, 156, 162 Weisbach, A., 172 Western hemisphere, meteorites of, Widmanstatten figures, 92, 127 Winchell, N. H., 171 Wohler, F., 124 Wright, A. W., 191, 193 Yale Collection, 223 Yearly falls, 37 Young, C. A., 217 INDEX OF METEORITES Adargas, 57, 5 Admire, 48, 165, 187 Agen, 48 Agram, see Hraschina Alais, 161, 189 Albacher Miihle, see Bitburg Albareto, 140 Aleppo, 48 Alexejewka, see Bachmut Alfianello, 105 Algoma, 68 Allegan, 139, 166, 195 Anderson, 187 Angra dos Reis, 109, 116, 124, 180, 184, 188 Arispe, 49 Aubres, 139 Aussun, 29, 176 Babb's Mill, 74, 203 Bachmut, 202 Bacubirito, 55, 58, 76, 91 Ballinoo, 150 Bandong, 49 Barbotan, 29, 46, 48 Barratta, 49 Bath Furnace, 66, 67 Beaconsfield, see Cranbourne Bear Creek, 146 Benares, 29 Bendego, 56, 58, 132, 145, 155, 156, 164, 172 Bethany, 48, 50, 95 Bingera, 98, 126 Bischtiibe, 49, 135, 154, 162, 168 Bishopville, 78, 85, no, 139, 161, 171, 173, 174, 176 Bitburg, 202 Bjurbole, 58, 83, 146, 166 Blansko, 49 Bluff, 85, 87, ill Bohumilitz, 154 Bolson di Mapimi, see Coahuila Boogaldi, 70, 71, 72, 73, 116 Borgo San Donino, 48 Borkut, 127 Braunau, 14, 20, 28, 49, 66, 149 Breitenbach, see Steinbach Bremervorde, 49 Brenham, 48, 101, 185, 186, 187 Bridgewater, 90 Bustee, 78, no, 138, 172, 174, 177, 178 Butsura, 46, 49, 70 Cabin Creek, 13, 28, 59, 62, 65, 79 Caille, 135, 143, 162 Cambria, 154 Canyon Diablo, 23, 48, 52, 70, 119, 120, 121, 126, 132, 135, 144, 154, 155, 156, 157, 160, 163, 166 Cape Girardeau, 175 Cape Iron, 159, 203 Cape York, 54, 57, 58, 220, 225 Carcote, 120, 152 Carlton, 66, 90, 91, 148 Carthage, 164 Casas Grandes, 57, 58, 135, 154, 225 Castalia, 46 Castine, 5 Chail, 48 Chantonnay, 85, 88 Charcas, 58, 135, 162 Charlotte, 70, 79, 192 Charwallas, 160 Chassigny, 108, 1 1 1 , 1 24, 1 8 1 , 1 84, 1 87, 190 Chateau Renard, 105 Chulafinnee, 123 Chupaderos, 49, 55, 134, 225 Clarac, see Aussun Cleveland, 66, 143 Coahuila, 48, 50, 97, 141, 154, 164, 166 Cold Bokkeveld, 46, 48, 161, 192 Collescipoli, 188 Concepcion, see Adargas Copiapo, 173, 174, 202 Cosby Creek, 49, 122, 123, 134, 135, 142, 146, 154, 162 Costilla, 66 Crab Orchard, 49, 102, in, 160, 162, 168 Cranbourne, 57, 58, 123, 135, 146, 154, 155, 162, 193 Cronstadt, 48 Cynthiana, 176 Danville, 146 Deep Springs, 159 Dehesa, 120 De Sotoville, see Tombigbee River Descubridora, 90, 91 Dhurmsala, 48, 105, 106, 194 231 232 INDEX OF METEORITES Dona Inez, 163 Drake Creek, 49 Durala, 70 Eagle Station, 183, 187 Elbogen, 154, 220 El Morito, 56, 58, 65, 225 Ensisheim, 165, 220 Estherville, 21,46,48, in, 127, 142, 176 Estacado, 195 Farmington, 21, 88, 89, 142, 190 Fisher, 171 Forest, 46, 48 Forsyth, 159 Fort Duncan, 98 Frankfort, 168 Futtehpore, 48 Gargantillo, see Tomatlan Gibbs, see Red River Glorieta, 49, 91, 132, 134, 151, 154, 162 Gnadenfrei, 176 Goalpara, 63, 126, 185, 188 Gopalpur, 127, 169 Grazac, 48 Great Nama Land, see Bethany Greenbrier County, 126, 165 Grossliebenthal, 169, 175 Griineberg, 142 Hainholz, no, in, 160, 176 Harrison County, 49 Hessle, 10, 46, 48, in, 169, 189, 190 Hex River Mts., 75, 98, 149, 155, 162 Holbrook, 46, 48, 49, 59 Holland's Store, 98, 149 Homestead, 16, 20, 46, 47, 48, 104, in, 192, 193 Honolulu, 125, 152 Hraschina, 10, 20, 49, 92, 154 Hvittis, 21, 139, 169, 174 Ibbenbiihren, 109, 171, 174 Ilimae, 134, 143, 148 Imilac, 48, 126, 1 86, 187, 202 Inca, 48 Indarch, 139 Indian Valley, 98, 149 Ivanpah, 162 amestown, 91 arnyschewa, see Pawlodar elica, 48, 146 ewell Hill, 143 onzac, 48, 64, 109, 167 uncal, 80 uvinas, 109, in, 140, 167, 168, 179, 180 Kendall County, 120, 154 Kenton County, 135 Kesen, 48 Khairpur, 9, 46, 48 Khetree, 48 Kilbourn, 22 Killeter, 48 Klein-Wenden, 166 Knyahinya, 15, 20, 23, 46, 48, 49, 58, 103, 213 Kodaikanal, 180, 187, 202 Kokstad, 75, 162 Krahenberg, 20 Krasnojarsk, 135, 182, 186, 187, 201 L'Aigle, 46, 48, 49, 81, 85, 161, 166, 222 Lance, 10, 49 Lancon, 88 Laurens County, 157 Lenarto, 191 Lick Creek, 98, 126, 159 Limerick, 46 Lime Creek, see Limestone Creek Limestone Creek, 155, 157, 160, 162 Linville, 203 Llano del Inca, 176, 177 Locust, 162 Lodran, 109, 165, 166, 172, 174, 182, 186 Long Island, 58, 66, 89 Losttown, 49 Luotolaks, 180 Macao, 29, 46, 48 Madrid, 48, 170 Magura, 58, 90, 120, 121, 123, 132, 133, 135, 142, 151, 154, 155, 156, 162, 192 Manbhoom, 48, in Manegaum, 109, no, 171, 172, 174 Marion, 49, 160 Marjalahti, 154, 165, 166, 187 Massing, 29, 172 Mazapil, 14, 28, 123, 208 Medwedewa, see Krasnojarsk Merceditas, 142 Mezo-Madarasz, 48, 103, 107, 127, 190 Middlesbrough, 21, 125 Mighei, 176 Mincy, 102, no, 183, 186 Misshof, 172 Misteca, 134, 135, 149, 162 Mocs, 7, 20, 46, 48, 49, 81, 82, 83, 85, 86, 172, 179, 194 Modoc, 48, 8 1 Molina, 175 Monte Milone, 48 Mordvinovka, 125 Morito, see El Morito Morristown, 139, 160, 168 INDEX OF METEORITES 233 Mount Joy, 92, 154 Mount Vernon, 166 Muchachos, see Tucson Murphy, 98 Nagaya, 160, 161, 189 Nakhla, 29, 177, 178 Narraburra, 116, 150 Nedagolla, 203 Nenntmannsdorf, 98, 146 Nerft, 125, 152 Ness County, 48 Netschaevo, 202 New Concord, 13, 46, 48, 191 N'Goureyma, 68, 203 Nocoleche, 146 Nowo Urei, 109, 118, 178, 180 Nulles, 48 Ochansk, see Tabory Orgueil, 20, 48, 51, 161, 166, 189 Orvinio, 10, 85, 87 Pallas, see Krasnojarsk Parnallee, 127, 190, 192 Pawlodar, 187 Penkarring Rock, see Youndegin Petersburg, 168, 180 Pillistfer, 29, 46 Ploschkowitz, 48 Prambanan, 79 Pnmitiva, 203 Pultusk, 20, 46, 48, 49, 85, 88, in, 161, 176, 192, 194, 213 Puquios, 90, 91 Quenggouk, 15 Quesa, 89 Quinn Canyon, 58, 65, 164 Rasgata, 162 Red River, 58, 93, 192 Renazzo, 105, m, 127 Richmond, 142, 176 Rittersgriin, see Steinbach Rochester, 52, 176 Roda, 124, 176 Rokicky, 187, 202 Rowton, 14, 146, 193 Sacramento, 89 St. Genevieve County, 160 St. Michel, 21, 83 Saintonge, see Jonzac Saline, 152 Salt Lake City, 175 San Emigdio, 176 San Gregorio, see El Morito Santa Rosa, 149, 155 Sao Juliao, 154, 159, 162 Sarepta, 151 Sauguis, 8 Schwetz, 162, 164 Scottsville, 98 Seelasgen, 146, 149, 151, 155 Segowlee, 48 Sewrukof, 166 Shalka, no, 124, 166, 171, 173, 174 Shergotty, 109, 163, 170, 179, 180 Shingle Springs, 192, 203 Siena, 48, in Sierra de Chaco, see Vaca Muerta Sierra di Deesa, see Copiapo Siratik, 203 Smith Mountain, 157 Smith ville, 49, 122 Sokobanja, 8, 48, 146, 176 Stalldalen 49, 85, 87, 176 Stannern 46, 48, 109, 112, 168, 180 Staunton, 49, 126, 132, 135, 143, 191 Steinbach, 49, 146, 162, 172, 174 Summit, 98 Tabory, 7, 48, 89, 126 Tadjera, 88, 176 Tazewell, 146, 150, 154, 157, 192 Tennasilm, 145, 169 Tieschitz, 127 Toluca, 48, 52, 76, 91, 99, 120, 123, 132, 134, 146, 148, 151, 154, 160, 161, 162, 163, 169, 195 Tomatlan, 48 Tombigbee River, 98, 154 Tomhannock Creek, in Tonganoxie, 126 Toulouse, 48 Trenton, 49, 143 Tucson, 49, 74, 81, 203, 225 Vaca Muerta, in, 146, 162, 168 Victoria West, 143, 202 Waconda, 169, 176 Warrenton, 12 Welland, 126, 132, 135, 136 Weston, n, 20, 46, 48, in, 192 Wichita County, 134, 135, 144, 155, 156 Willamette, 55, 58, 65, 225 Wold Cottage, 1 1 1 Youndegin, 49, 81, 122 Zabrodje, 169 Zacatecas-, 58, 154, 202, 225 Zavid, 184 Zsadany, 49, 152, 176 14 DAY USE RETURN TO DESK FROM WHICH BORROWED EARTH SCIENCES LIBRARY This book is due on the last date stamped below, or on the date to which renewed. Renewed books are subject to immediate recall. -rnw -""" wmw^ JUN 12 1973 r^zw u-Jssa-i. i..