yv_-ir_n_ REESE LIBRARY OF ! UNIVERSITY OF CALIFORNIA. No. ^// ^ 9 fa . C7.7S5 No. SPECTRUM ANALYSIS, BY JOHN LANDAUER, LL.D., > Member of the Imperial German Academy of Naturalists. AUTHORIZED ENGLISH EDITION BY J. BISHOP TINGLE, PH.D., F.C.S., Instructor of Chemistry in the Lewis Institute, Chicago, III. FIRST EDITION 1 . FIRST THOUSAND. NEW YORK: JOHN WILEY & SONS. LONDON: CHAPMAN & HALL, LIMITED. 1898. Copyright, 1898, BY J. BISHOP TINGLE. ROBERT DRUMMOND. ELECTROTYPES AND PRINTER, NEW YORK. AUTHOR'S PREFACE. THIS work originated as a reprint of an article on Spec- trum Analysis recently contributed to Fenling-Hell's " Neues Handworterbuch der Chemie "; it was published as a sepa- rate book at the request of a number of competent authorities, but not without some hesitation on the part of the author, because in treating a subject in an encyclopedic article regard must be paid to the whole plan and scope of the work, whilst in a separate book the author is quite independent. The favorable reception accorded to the book when published gives rise to the hope that shortcomings arising from its origin are to some extent counterbalanced by a fulness of contents brought together in small space, by the strictly historical treatment of the subject adopted through- out the book, by tolerably full bibliographical references, and by the care which has been bestowed on the numerical tables serving for reference. In order to secure a degree of uni- formity hitherto wanting, the older measurements have been recalculated so as to bring them into accord with Rowland's system of wave-lengths. THE AUTHOR. BRAUNSCHWEIG, 1898. iii ABBREVIATIONS. THE following abbreviations have been used in the biblio- graphical references: A. B. A. = Abhandlungen der Koniglichen Akademie der VVissenschaften zu Berlin. A. c. p. = Annales de chimie et de physique. B. A. R. = British Association Reports. C. N. = Chemical News. C. r. = Comptes rend us hebdomadaires des Seances de 1' Academic de Sciences, Paris. N. = Nature. N. A. S. U = Nova Acta Regiae Societatis Scientiarum Upsaliensis. p. A. = Poggendorff's Annalen der Physik und Chemie. P. M. = Philosophical Magazine. P. M. P. S. = Proceedings of the Manchester Literary and Philosophical Society. P. R. S. = Proceedings of the Royal Society. P. R. S. E. Proceedings of the Royal Society, Edin- burgh. P. T. = Philosophical Transactions. P. T. E. = Philosophical Transactions, Edinburgh. T. R. S. E. = Transactions of the Royal Society, Edin- burgh. W. A. = Wiedemann's Annalen der Physik und Chemie. TRANSLATOR'S PREFACE, THE claim of spectrum analysis to a place in a chemical curriculum is steadily obtaining increased recognition, and its importance is generally admitted both for students preparing for teaching, and for those who wish to engage in technologi- cal work. The subject may rightly demand a wider field since its pursuit furnishes so many opportunities for an excellent training in accuracy of observation and manipulative skill that it might, with great advantage, find place in a general science course. The expense is by no means prohibitive, and is almost entirely confined to the first cost of the instruments, which, with proper care, last for years, and even with the cheaper and smaller ones, such as Browning's "Students' Spectroscope," which costs about $30, much interesting work can be done and valuable discipline obtained. Of the works on spectrum analysis hitherto published in English, none are suitable as text-books, either on account of their size and consequent cost, or from the manner in which the subject is presented. It is hoped that this little book may, in some degree, sup- ply this lack. There has been no attempt to treat the subject exhaus- tively, but rather to indicate the more salient points of theory, etc., leaving it to the teacher to complete and expand them at his own discretion. No doubt it would be well if all students were compelled to take a course in general physics before attacking chemistry, VI 11 TRANSLATOR'S PREFACE. but at present this is a state of things not realized in practice ; to those who have followed such a course the physical section of the book should be superfluous ; but it may serve to call the attention of the others to matters on which they should obtain more instruction. The tables of wave-lengths will, it is hoped, be useful in the practical work which would prob- ably constitute the greater portion of the course. The posi- tion of the more prominent lines and bands can, by their help, be at once ascertained, and their actual occurrence and identification facilitated. LEWIS INSTITUTE, CHICAGO, ILL., Dec. 1897. CONTENTS. CHAPTER I. PAGE INTRODUCTORY HISTORICAL . t i Introductory, i. Historical, 2. Bibliography of Works on Spectrum Analysis, 8. CHAPTER II. PHYSICAL PROPERTIES OF LIGHT n Wave-length, n. Reflection; Refraction, 12. Prisms, 13. Dis- persion, 14. Abnormal Dispersion, 15. Pure Spectra; Gratings, 16. Diffraction, 17. Comparison of Diffraction and Refraction Spectra, 19. CHAPTER III. SPECTROSCOPES 20 Spectroscopes with Angular Vision, 20. Direct-vision Spectro- scopes, 28. Grating Spectroscopes, 31. CHAPTER IV. SPECTROSCOPIC INSTRUMENTS FOR SPECIAL PURPOSES . . . .35 Spectrometer, 35. Kriiss' Universal Spectroscope, 38. Spec- trophotometer, 39. Sorby's Microspectroscope, 40. Solar and Stellar Spectroscopes, 41. Stellar Spectrometers; Spectrographs, 43. Rowland's Concave Grating Spectrograph, 45. CHAPTER V. SPECTROSCOPIC ADJUNCTS . . 5* Flame Spectra, 51. Electric Arc, 54. Electric Spark, 55. Geiss- ler or Plucker's (Vacuum) Tubes, 58. Observation of the Invisible Regions of the Spectrum; Ultra-violet, 60. Infra-red, 61. Obser- vation of Absorption Spectra, 63. Measuring Appliances and Scales, 65. Drawings of Spectra, 68. X CONTENTS. CHAPTER VI. PAGE EMISSION-SPECTRA .69 Line Spectra, 71. Influence of Temperature and Pressure, 72. Lockyer's Long and Short Lines, 73. Influence of Magnetic Current, 75. Absorption Spectra; Kirchhoffs Law, 76. Influence of the Temperature and the Physical State of Substances, 77. Influence of the Solvent, 78. Influence of the Optical Density; Fluorescence and Absorption, 79. Relationship between the Lines of an Element, 80. Relationship between the Spectra of Differ- ent Elements, 87. CHAPTER VII. SPECTRA OF THE ELEMENTS 9 2 Unit of Measurement, 92. Scale of Colors ; Delicacy of Spectrum Reactions, 95. Spectra of the Elements (in Alphabetical Order), 96. CHAPTER VIII. ABSORPTION SPECTRA 174 Absorption by Gases and Liquids, 175. Different Salts of the same Colored Base or Acid, 177. Relationship between Molec- ular Structure and Absorption Spectrum, 178. Absorption in the Visible Portion of the Spectrum, 179. Absorption in the Ultra- violet, 182. Absorption in the Infra-red, 184. CHAPTER IX. THE SOLAR SPECTRUM .......... 186 The Fraunhofer Lines, 186. The Chemical Composition of the Sun, 190. Rowland's Table of Wave-lengths of the Fraunhofer Lines, 191. Telluric Lines of the Solar Spectrum, 201. Limits of the Investigation, 202. Physical Condition of the Sun, 203. Solar Nucleus ; Photosphere ; Sun-Spots. 204. Solar Faculae ; Reversing Layer ; Chromosphere and Prominences, 205. Corona, 207. CHAPTER X. OTHER CELESTIAL BODIES . . .208 Fixed Stars, 208. Planets and Moon, 209. Comets; Meteor- ites and Shooting Stars ; Nebulae, 210. Aurora Borealis; Zodiacal Light; Lightning, 211. Displacement of the Lines, 212. SPECTRUM ANALYSIS, CHAPTER I. INTRODUCTORY. HISTORICAL. SPECTRUM ANALYSIS is a chemico-analytical method by means of which it is possible to determine the constituents of a substance, by observing the refraction (dispersion), or the diffraction of light-rays. Its further development offers a means of investigating the molecular structure of matter. The image which is produced when light-rays are refracted is termed a spectrum. White-hot solid bodies emit rays of all refrangibility, and give a continuous spectrum; glowing gases or vapors emit rays of definite refrangibility, and therefore yield a discontinuous spectrum consisting of bright lines which are characteristic of each substance, and which consequently serve for its identification whether it occurs alone, or together with other bodies. When the rays from a white-hot solid pass through a colored medium some of them are retained giving an absorption spectrum, which varies with the chemical composition of the medium. Spectra-reactions are characterized by an extreme delicacy far exceeding that of chemical tests, and therefore their employment has led to the discovery of a number of new elements which occur only in small quantity. Since the distance of the source of light has little effect on a spectrum, the method can be employed for the investigation of celestial 2 SPECTR UM ANAL YSIS. bodies: it has extended our knowledge of .their nature to an extent which was previously entirely unattainable. Historical. 1 Spectrum Analysis was founded by Kirch- hoff and Bunsen in 1859, and subsequently developed. Other observers had previously noticed spectrum lines, and had suggested the application of spectroscopic observations to chemical analysis, but their efforts were fruitless, as at that time it was not certain whether the bright lines of a glowing gas were solely dependent on its chemical composi- tion. The sodium reaction was particularly misleading as it was often observed when the presence of this metal was not suspected, and was therefore variously ascribed to sodium, to sulphur, or to water. The yellow sodium flame was first noticed by Thomas Melville 2 in 1752, but he was unable to determine its origin. John HerscheP in 1822 investigated the spectra of many colored flames, particularly those given by strontium, copper, and boric acid, and in 1827 showed that by this means the substances giving the colors could be recognized even when present only in extremely small quan- tity. Fox Talbot 4 in 1826 expressed himself still more definitely, stating that if his theory that certain bodies gave characteristic lines should prove to be correct, then a glance at the prismatic spectrum of a flame would suffice to. identify substances which would otherwise require a tedious chemical analysis for their detection. In 1834 he correctly described the lithium and strontium spectra, and again pointed out that 1 Kopp, Entwickelung der Chemie in der neuren Zeit (Miinchen, 1873), pp. 215, 642. Kirchhoff, Zur Geschichte der Spectralanalyse. P. A. 118, 94, 102. Brewster, C. r. 62, 17. Kahlbaum, Aus der Vorgeschichte der Spectralanalyse (Basel, 1888). Rosenberger, Geschichte der Physik, 3 (Braunschweig, 1890). Stokes, N. 13, 188. P. M. [4] 25, 250. Talbot, P. R. S. E. 7, 461. 2 Edinb. Phys. and Lit. Essays, 2, 12. 8 T. R. S. E. (1823) 9. P. A. (1829) 16. On the theory of light (Lon- don, 1828) 4 Brewster's Journ. of Sci. 5. P. M. (1833) [3] 3, 35; (1836), 9, 3. IN TROD UCTOR Y. HIS TO RICA L. 5 such optical methods permitted of the identification of these elements with a minimum quantity of substance, and with an exactitude equalling, if not excelling, that attained by any other process. Doubt was, however, cast on this conclusion by contradictory statements in the same communications, and the method of analysis was rendered fundamentally dubious, because, in opposition to Herschel, Talbot maintained that the reactions could be produced by the simple presence of the substance in the flame, its volatilization not being necessary. W. A. Miller 1 published in 1845 an investigation on trie spectra of the alkali metals; diagrams were given, but the results did not constitute any great advance, as he had employed a luminous flame, and was therefore unable to determine what was characteristic of any particular metal. In 1856 Swan 3 definitely proved that the yellow line which is almost always present is peculiar to sodium compounds, and that the frequency of its occurrence is due to the almost universal distribution of sodium salts. In his work on the prismatic spectra of the hydrocarbons Swan showed that the lines observed are constant in position; he thus made a valu- able contribution towards the solution of the question as to whether the bright; lines of a glowing gas are exclusively dependent on its chemical composition. The de6nite and general answer to this problem was, however, not given by Swan, but by Kirchhoff and Bunsen. The spectra of the electric spark had been under obser- vation simultaneously with those of flames; Wollaston 3 detected a large number of bright lines, but without offering any clue to their origin. He was also the first to describe the dark lines in the solar spectrum, and he improved the apparatus employed by substituting a narrow slit for the cir- cular opening which Newton had used to admit the light. 1 B. A. R. 1845. P. M. [3] 27, Ci. 2 T. R. S. E. 3, 376; (1857)21,353. 3 P. T. 1802 p. 365. 4 SPECTRUM ANALYSIS. Fraunhofer 1 was scarcely more successful than Wollaston so far as the origin of the bright lines was concerned; his fame rests on the discovery of the diffraction grating, the measure- ment of wave-lengths which its use permitted, and on the observation of the dark lines in the solar spectrum which bear his name. He drew 350 of these, and finding that they varied from those observed in stellar spectra, he concluded that they originate in the sun and stars, and are not due to the earth's atmosphere. Wheatstone 3 in 1835 found that with the use of different metallic electrodes the spectra vary, but they remain constant no matter whether the discharge takes place in air, hydrogen, or in a vacuum; he therefore concluded that the metal is volatilized, but not burnt, by the passage of the spark. He published drawings of the spectra of sodium, mercury, zinc, cadmium, bismuth, tin, and lead, and recommended the method for analytical purposes. The spectra of various metals volatilized in air were studied, although less thoroughly, by Foucault 3 in 1849; ne also observed the dark /Mine, since known as the reversed sodium line, but failed to draw the important conclusion from this which Kirchhoff subsequently made. Masson, 4 who improved the method of working, using condensers charged by induction-currents, investigated the spark-spectra of iron, tin, antimony, bismuth, copper, lead, cadmium, and carbon; in all these cases he noticed that the lines due to moist air were present, although he was ignorant of their origin. This was indicated by Angstrom's & important work published in -1853. He showed that the lines which occur in the space between the electrodes are due to air or to any other gas 1 Denkschriften der Miinchener Akad., 1814, 1815; Gilbert's Ann. 74, 337- 2 B. A. R. 1835. C. N. 3, 198. P. M. [3] 7. 3 Institut. 1849, p. 44. * A. c. p. (1851) [3] 31, 295. b K. Vetenskaps Akad. Handl. (Stockholm, 1853), P- 335- P. A. (1855) 94, 141. INTRODUCTORY. HISTORICAL. 5 which may be present, whilst those close to the electrodes are given by the metals. Angstrom also drew and described the spectra of a large number of metals and non-metals, and almost discovered the relationship between the emission and absorption of light, since he stated, in accordance with a sug- gestion of Euler, that at a common temperature bodies absorb the same vibrations which they are capable of producing. In 1858 Plucker 1 commenced his investigations of the spectra produced by the passage of an electric current through highly rarefied gases. He found that the elementary gases, or the constituents liberated from compound gases, are char- acterized by bright lines. Similar work was pursued by van der Willigen, 2 who in 1859 also showed that platinum electrodes moistened with a salt solution give the spectrum of the salt, and that it is therefore unnecessary to use the metal itself in order to obtain its spectrum. In the same year Kirchhoff and Bunsen 3 published their work " Chemische Analyse durch Spectralbeobachtungen" ; their results were obtained to some extent independently of previous investiga- tors who, whilst frequently on the right path, had failed to reach the goal. They reduced spectral phenomena to a chemical-analytical method, and definitely proved that the bright lines produced by a glowing gas are dependent only on its chemical composition. This law still forms the basis of spectrum analysis, but their second proposition has been subsequently considerably modified; it states that the manner in which the constituents of a substance are combined is with- out influence on their spectra, and that these are also almost entirely unaffected by the temperature and pressure of the vapor. After Roscoe and Clifton 4 had called attention to the difference between the spectrum of an element and those of its compounds, A. Mitscherlich * showed, in 1863, that every 1 P. A. 103, 88; 104, 113, 622; 105, 67; 107, 77, 415. ' l P. A. 106, 610; (1859) 107, 473. 3 P A. 110, 167. 4 P. M. P. S. 1862. 5 P. A. (1863) 121, 3. "r ^T TTT- T-I T- ri T m-7- 6 SPECTRUM ANALYSIS. compound has its own peculiar spectrum, and that the exhi- bition of identical spectra by the various salts of an element is caused by these undergoing dissociation. In their first communication Kirchhoff and Bunsen described the spectra of the metals of the alkalis and alkaline earths, and showed the great delicacy of the method, which permits of the recog- nition of substances when present in quantity far too small for detection by the ordinary processes; they also pointed out the great extension which it gives to our knowledge of the distribution of the elements, and indicated that it would probably lead to the recognition of new ones. The correct- ness of this view has been proved by the discovery of caesium, rubidium, thallium, indium, gallium, and many metals of the rare earths, all by means of spectrum analysis. The development of spectrum analysis received a special impulse from its application to astronomy. Kirchhoff ' proved mathematically that for every ray of light the relationship between the emissive and absorptive powers of all bodies is. alike at uniform temperatures; this explained the origin of the Fraunhofer lines, and led to the investigation of the chemical composition of the sun and its atmosphere. The discovery of this law of exchanges induced Kirchhoff to prepare more exact drawings of the solar spectrum, and to accurately com- pare the positions of the Fraunhofer lines with those in the spectra of many terrestrial substances. He employed for this purpose an arbitrary scale, as did also Huggins, 2 who extended these observations. To Angstrom 3 belongs the credit of sub- stituting the wave-length for the scale as a means of determin- ing the position of the lines, and his measurements, and atlas of the solar lines, remained for twenty years the foundation of all spectroscopic investigations. Angstrom's work was 1 Monatsber. Berl. Akad., Oct. 27, 1859. 2 P. T. (1864) 154. 3 Recherches sur le spectre normal du soleil avec atlas de 6 planches, Upsala, 1868. IN TROD UCTOR F. HIS TO RICA L. 7 confined to the visible portion of the spectrum; it was com- pleted by Cornu's * researches on the ultra-violet, and by Langley's 2 and Abney's 3 on the infra-red. After Angstrom's death, Thalen 4 showed that the metre he had employed was incorrect, and that consequently his wave-length determina- tions were too small. This was confirmed by Miiller and Kempf 5 in 1886; their measurements of 300 solar lines were carried out with great care, and became the basis of the Potsdam system. All these determinations were, however, exceeded in accuracy by Rowland's 8 Atlas of the solar spec- trum, and his reproductions of normal lines, published in 1888 and 1893 respectively. His discovery of the concave grating in 1891, and his " coincidence " method of determining the relative position of lines, has greatly aided spectroscopic work, since it admits of the production of photographs without the use of a lens, thus insuring a high degree of comparative accuracy. For a considerable time the measurement of the spectra of terrestrial substances did not keep pace with that of the solar spectrum; Kirchhoff's and Huggins' determinations were duly superseded by the more accurate ones of Thalen, 7 but these were confined to the visible spectrum. Apart from W. A. Miller's 8 incomplete work on the ultra-violet in 1862, Lockyer 9 in 1881 was the first to accurately investigate the 1 Spectre normal du soleil. Partie ultraviolette (Paris, 1881), p. 22. 2 R M. [5] 21, 394; 22, 149; 26, 505. 3 P. T. 1880, p. 653. W. A. Beibl. 4, 375; 5, 507. C. r. 90, 182. 4 Spectre du fer. Acta R. Soc. Sclent. Upsala, (1884) [3], p. 49. W. A. Beibl. .9, 520. 5 Publ. d. Astrophys. Obs. zu Potsdam (1886), 5. 6 Photographic Map of the normal Solar Spectrum, Johns Hopkins Univ., Baltimore. Astronomy and Astrophysics (1893), 12, 321. P. M. (1894) [5] 36, 49. 7 N. A. $. U. (1868) [3] 6. 8 P. T. (1862) 152, 861. 9 P. T. (1873) 163, 253, 639; (1874) 164, 479, 805. P. R. S. 25, 546; 27, 49, 279, 409; 28, 157. 8 SPECTRUM ANALYSIS. subject, but he soon quitted it, and its fuller examination was reserved for Hartley and Adeney, 1 and Liveing and Dewar." Since 1888 Kayser and Runge 3 have met with great suc- cess in their important task of measuring the emission-spectra of terrestrial substances by Rowland's method. They com- menced the work in order to determine the relationship of the various lines of an element, and also that of the lines of different elements. Attempts had been made in this direction shortly after the discovery of spectrum analysis by Kirchhoff and Bunsen; it was at first believed that the relationship of the lines was similar to the sound-waves of a vibrating string, which consist of a fundamental note and harmonic overtones. This view was shown by Schuster 4 in 1880 to be incorrect, and in 1885 Balmer 5 discovered a formula which accurately reproduces the hydrogen lines in wave-lengths. These inves- tigations, together with the observations of Liveing and Dewar 6 on harmonic series of similar lines, are naturally con- nected with Kayser and Runge's work, which has led to the discovery of the methodical structure of a series of spectra. Rydberg, 7 working independently of Kayser and Runge, has obtained similar results. Investigations of this nature have tended to greatly widen the domain of spectrum analysis. BIBLIOGRAPHY OF WORKS ON SPECTRUM ANALYSIS. CAPRON. Photographed Spectra. 1877. CAZIN. La spectroscopie. Paris, 1878. DEMARgAY. Spectres electriques. Paris, 1895. DlBBlTS. De Spectraal-Analyse. Rotterdam, 1869. P. T. 1884, p. 63. P. T. 174. 187. P. R. S. 34, 119, 123. W. A. Beibl. 6, 934; 7, 849. A. B. A. 1888-1894. Runge B. A. R. 1888, 576. B. A. R. 1880. W. A. (1885)25. P. T. (1883) 174, 208. C. r. (1890) 110, 394. K. Vetenskaps Akad. Handl., 23 (Stockholm, 1890). INTRODUCTORY. HISTORICAL. 9 DIETERICI. Spectralanalyse in Ladenburg's Handworter- buch der Chemie. Breslau, 1892. DRAPER. Catalogue of Stellar Spectra. Cambridge, 1895. GANGE. Die Spectralanalyse. Leipzig, 1893. DE GRAMONT. Analyse spectrale directe des mineraux. Paris, 1895. GRANDEAU. Instruction pratique sur 1'analyse spectrale. Paris, 1863. HiGGS. Photographic Atlas of the Normal Solar Spectrum, 1894. HUGGINS. Results of Spectrum Analysis applied to the Heavenly Bodies. London, 1870. KAYSER. Lehrbuch der Spectralanalyse. Berlin, 1883. (Contains measurements of spectra and a very complete review of the literature.) Spectralanalyse in Winkelmann's Handbuch der Physik (Encyclopedic der Naturw.). Breslau, 1894. KLINKERFUES. Die Spectralanalyse und ihre Anwendung in der Astronomic. Berlin, 1879. V. KOVESLIGETHY. Grundzuge einer theoretischen Spectral- analyse. Leipzig, 1890. KONKOLY. Handb. der Spectroskopiker. Halle a. S., 1890. G. AND H. KRUSS. Colorimetrie und quant. Spectralanalyse. Hamburg and Leipzig, 1891. LECOQ DE BoiSBAUDRAN. Spectres lumineux. Paris, 1874. LlELEGG. Die Spectralanalyse. Weimar, 1867. LOCKYER. The Spectroscope and its Use. London, 1874. - Studies in Spectrum Analysis. New York and Lon- don. 1878, LORSCHEID. Die Spectralanalyse. Munster, 1875. MACMUNN. The Spectroscope. London, 1888. PROCTOR. The Spectroscope. London. ROSCOE. Spectrum Analysis. Fourth edition, revised by the author and A. Schuster. London, 1885. (Con- tains popular lectures on the subject supplemented by 10 SPECTRUM ANALYSIS. extracts from the more important original memoirs, and a good bibliography.) SALET. Traite elementaire de spectroscopie, I. Fascicule. Paris, 1888. SCHEINER. Die Spectralanalyse der Gestirne. Leipzig, 1890. (Contains a comprehensive bibliography.) SCHELLEN. Die Spectralanalyse in ihrer Anwendung auf die Stoffe der Erde und die Natur der Himmelskorper. Braunschweig, 1883. English translation by J. and C. Lassel edited by W. Huggins. New York, 1872. SECCHI. Die Sonne, German by Schellen. Braunschweig, 1872. THALEN. Spectralanalyse expose och Historick, med en Spectralkarta. Upsala, 1866. TUCKERMAN. Index to the Literature of the Spectroscope. Washington, 1888. VlERORDT. Anwendung des Spectralapparates zur Photo- metric und zur quant. Analyse. Tubingen, 1873. VOGEL H. W. Prakt. Spectralanalyse irdischer Stoffe. Berlin, 1889. (Deals chiefly with practical analysis, and particularly with absorption-spectra.) WATTS. Index of Spectra. Manchester, 1889. (Contains complete measurements of spectra and a very full bibliography. Supplements to the index appeared in the B. A. R., London.) YOUNG. The Sun. New York and London, 1897. See also text-books on physics, amongst others: A. Lommel, Lehrb. d. Experimentalphysik (Leipzig, 1893); Muller-Pouillet's Lehrb. d. Physik, 9. Aufl. v. Pfaundler (Braunschweig, 1894); Winkelmann, Handb. d. Physik (Breslau, 1893); Kelvin & Tait, Elements of Natural Philos- ophy; Tait, Light; Tyndall, On Light; Wright, Light. CHAPTER II. PHYSICAL PROPERTIES OF LIGHT. 1 ACCORDING to Huygens' universally accepted theory, light consists of wave-motions of the ether, the vibrations being transmitted from particle to particle with an extremely high velocity in straight lines; the vibrations of the particles of ether are at right angles to the path of the ray. On account of the great elasticity of the ether, and the ease with which the vibrations are further propagated, single rays can- not be obtained, but only pencils consisting of a number of rays, which may be considered to be parallel if it is assumed that the vibrations are very small, or at a great distance from the source. The varying frequency of the vibrations produces in the eye the effect of color; the number of vibrations is constant for each color, but in a given medium the wave- length differs. Since all light-rays are transmitted with a uniform velocity in the free ether or in a vacuum, and almost so in air, the number of vibrations is small or great in propor- tion as the waves are long or short. Wave-length. It is possible to directly determine the wave-length corresponding with a given color in air, and it is found that at the extremity of the visible red the wave-length (A) of the ^4 -line = 0.00076 mm., that of the yellow ^,-line = 0.000589 mm., and that of the AT-line at the limit of the visible violet = 0.00039 mm. The velocity (v) of light is 1 Comp. Fehling-Hell's Handworterbuch, 4, 87, and text-books of Phy- sics. 12 SPECTRUM ANALYSIS. known to be about 300,000 kilometres per second; the num- ber of vibrations () is obtained by the expression n = In this manner it is found that the number of vibrations of the above three lines = 395, 509, and 763 billions per second respectively. These numbers are inconceivably great, and awkward to write, and it is therefore usual to define the color by the wave-length, although this varies with the medium. In dealing with wave-lengths measured in a vacuum, the millionth part of a millimetre = o.ooi mikron is taken as the unit, and, in accordance with Kayser's suggestion, it is repre- sented by the symbol /*/*; a tenth part of this = o. !//// is termed an Angstrom's unit. Reflection. The light which falls on a rough nonlumin- ous body is partly absorbed or transmitted and the remainder thrown back on all sides, thus making the object visible; but smooth polished surfaces mirrors only reflect the light in certain definite directions; the perpendicular produced at the point of the reflecting surface where the ray impinges is in the same plane as the incident and reflected ray, and both form identical angles with it. Refraction. The passage of light from one medium to another for instance, from air into water or glass causes a part of it to be thrown back in accordance with the law of reflection, whilst the remainder traverses the new medium, but not in a straight line; its path is deflected (Fig. i), a process which is known as refraction. The incident ray ae and the refracted ray eb are in the same plane as trie normal ed of the new medium; if the light passes from a rare to a denser medium, the FlG - * refracted ray approaches the perpen- dicular, otherwise it recedes from it. In order that refraction may take place the incident ray must form an acute angle PHYSICAL PROPERTIES OF LIGHT. 13 with the normal; if it forms a right angle, it traverses the medium in a straight line. Every incident jjngle corresponds with a particular refractive angle; the sine ac of the incident angle i bears a definite relationship ;/ to the sine bd of the angle of refraction r, in accordance with Snell's law, so that sin i sin i n = or sin r = or sin i = n sin r. sin r n This relationship is termed the refractive index, or coeffi- cient of refraction, and differs for every transparent substance. The refractive indices of the following substances are for /Might, and at a temperature of 20: Water 1.3333 Alcohol , 1.3616 Carbon bisulphide 1 .6276 tf-Bromonaphthalene 1.6582 Ethylic cinnamate (at 18.8) 1.5607 Crown glass 1.5 15-1.615 Flint glass 1.614-1.762 Jena, heaviest silicate flint glass, No. 557 1.9625 Quartz (ordinary ray) 1 .5442 Fluor-spar 1.4339 Air (o and 760 mm.) 1.0002922 Plates with plane parallel surfaces cause the incident ray to be as much deflected towards the perpendicular as the issu- ing ray is bent from it; the two rays are therefore parallel to one another. Prisms. A prism is a wedge-shaped transparent object with two polished surfaces forming an angle with one another. The section of an ordinary simple glass prism forms an equi- lateral triangle (Fig. 2); the polished sides AB and AC are the refracting surfaces, enclosing the refracting angle a and 14 SPECTRUM ANALYSIS. forming the refracting edge A. If a ray of light in the plane of the section falls on one of the sides AB or AC, it is bent at its entrance and exit, in ac- cordance with the law of refrac- tion, and approaches the thick portion of the prism. The ex- tent of this deflection is equal to the angle d which the incident FIG. a. . r i and issuing rays torm with one another, and is also equal to the sum of the angle of incidence and that of the issuing ray, minus the refractive angle. The angles /', r, and a bear a certain interrelationship, and it is possible to calculate in what position of the prism the refrac- tion will be smallest; this can be confirmed by direct observa- tion. This minimum deviation occurs when the ray forms the same angle with the refracting surfaces externally and inter- nally, or, in other words, when it traverses the prism sym- metrically. The refractive index n of the material of which the prism is composed may be calculated by the expression _ sin \(d -\- a) fir ' " . 1 sm \a the minimum deviation and angle of refraction being measured by means of the goniometer. Dispersion. Refraction not only changes the direction of a ray of light, but, if it is not homogeneous, its nature is also modified; a ray of white light is converted into a rain- bow-colored band, as may be easily seen by the help of a prism. The polychromatic rays composing white light are transmitted with uniform velocity in a vacuum, but in a denser medium the more rapidly vibrating violet rays undergo a greater retardation than the red rays, which vibrate more slowly; the former are therefore refracted more strongly than PHYSICAL PROPERTIES OF LIGHT. 15 the latter. The component rays are more strongly refracted by passage through a second prism, but do not undergo any further decomposition; they are therefore simple or homo- geneous, and if combined by means of a lens white light is reproduced. These experiments, which are of fundamental importance both for spectrum analysis and for the theory of light, were first performed by Newton in 1668, and described in his 44 Opticks " in 1675. He allowed a ray of sunlight to pass through a small hole in a window-shutter into a darkened FIG. 3. room XY (Fig. 3); he passed the rays through a prism ABC, which caused them to be deflected and resolved into the colored band PT, which he termed a spectrum, and which was received on the white screen MN. The colored rays when viewed by Newton through a second prism gave the impression of white light, but when they were made to traverse it separately they were not further decomposed, but only underwent a second refraction. Abnormal Dispersion. The refractive index of a medium is, as a rule, greater the smaller the wave-length of the par- ticular light; in the visible spectrum the index steadily increases in passing from red to blue. Certain substances do not conform to this rule, their solutions, when employed as refracting and dispersing agents, exhibit the inverse relation- 1 6 SPECTRUM ANALYSIS. ship between refractive index and dispersion. This phenome- non is termed abnormal dispersion. 1 Pure Spectra. The colors obtained if the light is ad- mitted to the prism through a round opening, as in Newton's experiment, are never completely separated from one another, as the circular shape of the images causes them to overlap. In order to separate the various colors as completely as possi- ble, and obtain a pure spectrum, a narrow longitudinal slit has, since Wollaston's 2 time, been generally employed for the admission of the light. The number of images of the slit produced is equal to that of the different wave-lengths in the light employed, and consequently the narrower the slit the less do the images superpose; the spectrum thus obtained may be magnified to any desired extent. The resolving power of a prism, or system of prisms, is partly dependent on its dis- persion, but to a greater extent, as Rayleigh 3 has shown, on the distance which the ray traverses in its route through them. The thickness of heavy flint glass required to separate the ZMines 1.02 cm.; taking this value, roughly i cm., as unit, the resolving power of a prism of similar glass in the region of the ZMine is equal to its thickness in centimetres; in other parts of the spectrum it is inversely proportional to the cube of the wave-length, so that it is eight times greater in the violet than in the red, and therefore corresponds with the total thickness, or with the length of the base of the prism or system, but is independent of the number of prisms, of their angle, or of the order in which the various members of the system are arranged. Gratings. An optical grating consists of a plate with a large number of parallel lines ruled upon it ; this arrangement, 1 Christiansen, P. A. (1870) 141, 479; (1871) 143, 250. Kundt, P. A. (1871) 143, 259; 144, 128; (1872) 145, 67. Sieben, W. A. (1879) 8, 137. Sellmeier, P. A. 145, 396, 520; 147, 386, 525. H. v. Helmholtz, Monats- ber. Berl. Akad. (1874), p. 667. P. A. (1874) 154, 582. 2 P. T. (1802) p. 378. P. M. (1879) [5] 9, 269. PHYSICAL PROPERTIES OF LIGHT. If like a prism, produces spectra. Gratings were first employed by Fraunhofer, 1 who at first used a wire grating prepared by- winding thin wire over two similar screws of very fine thread, placed parallel to one another; later he engraved numerous fine lines, closely adjacent and at regular intervals, on gold- leaf backed by glass, and finally employed glass plates with opaque lines cut by means of a diamond. The preparation of gratings has been greatly improved in more recent times by the use of good dividing-machines. Two kinds of gratings are made, the transparent ones of glass, with as many as 800 lines per mm., and reflection gratings of speculum metal which reflects instead of transmitting the light; the latter are preferable for spectroscopic work, as less light is absorbed. Rutherfurd, in the United States, considerably improved the construction of the reflection grating, and since 1882 their preparation has been carried to an extraordinarily high degree of perfection by Rowland at Baltimore. His plane and con- cave gratings with 10,000, 14,438, and 20,000 lines per inch are almost faultless, and comparatively free from scratches caused by irregularity of the diamond-point (Ghosts). Diffraction. When a narrow illuminated slit is viewed through a glass grating, the lines of which are parallel with the edges of the slit, a bright image of it is observed with a series of spectra on each side; the violet rays with the shortest wave-length are nearest, and the red rays most distant, from the centre in each spectrum, and, if the colors are almost equally dispersed, the yellow will be found in the middle. The spectra are distinguished as of the first, second, . . . w/th order, counting from the centre. The spectra of the first order only are pure; the others are modified by the superpos- ing of other spectra, but they may be separated by means of a ^mall prism as described in the following chapter. The intensity of the illumination diminishes with ascending order of the spectra. 1 Denkschriften d. Miinchener Akad. (1822) 8. Gilbert's Ann. 74, 337. 1 8 SPECTRUM ANALYSIS. The production of spectra by means of gratings is due to diffraction; part of the light traversing the spaces between the rulings continues in a straight line, but a portion is bent sideways, or refracted, by the sharp edges of the opaque parts. The explanation of this phenomenon afforded by the wave- theory of light is as follows: The light-waves which fall on a fine slit cause the particles of ether present to vibrate; this motion is communicated to the neighboring particles and pro- duces an equal number of light-waves which reinforce, weaken, or neutralize each other, in accordance with the law of interfer- ence. The neutralization occurs in all directions in which the difference between two sets of waves is other than a whole wave-length. In the case of white light, diffracted by means of a grating, the image of the slit, in the middle, is white because at this point all the colors are superposed, but the colored rays which differ in phase by one wave-length collect at each side according to their wave-lengths, and form a spectrum of the first order; those rays with a greater differ- ence of phase forming the spectra of the second, third, . . . mth order. The wave-length may be determined, if the distance between the lines of the grating is known, by measuring the angle of diffraction with a gonimeter. Angstrom in this manner found the following values, in ten millionths of a milli- metre, for the Fraunhofer lines given: A B C D, E F G H 7604 6867 6563 5895 5269 4861 4307 3968 The dispersive power of a grating is dependent on the total number of the spaces into which it is divided, and on the order of the spectrum; in one of the first order 1000 lines per inch are necessary to separate the ZMines, whilst a large Rowland grating, in a spectrum of the first order, is capable of dividing two lines differing in wave-length by only 0.05 of an Angstrom. UNIVERSITY PHYSICAL PROPERTIES OF LIGHT. 1 9 Comparison of the Diffraction and Refraction Spectra. Diffraction spectra differ from those produced by refraction in the dispersion of the rays being proportional to the wave- length, and this uniform extension applies although the dis- persion increases with the number of lines on the grating. The refraction in the case of a prism spectrum increases with diminishing wave-lengths; the violet and blue rays are there- fore comparatively widely separated, and the red ones gath- ered together. The length of the spectrum is also influenced by the composition of the prism, so that results obtained with different spectroscopes are not directly comparable; diffractive spectra are therefore taken as typical or normal, and all scale readings with a prism spectroscope are reduced to wave- lengths. The prism spectrum has the advantage over the diffraction spectrum of greater brightness, only a small pro- portion of the light is lost by reflection and absorption, whereas with the grating a portion of the light passes through without being diffracted, a portion is weakened by interfer- ence, and the remainder is divided amongst a number of spectra instead of being concentrated into one as in the case of the prism. The prism spectrum is therefore employed where the illumination is comparatively feeble, the grating being used for intense light and in cases where a high dis- persion is necessary. A large Rowland grating in the region of the ZMine produces the same effect as a prism 126 cm. in thickness; in the violet this proportion changes in favor of the prism ; at A 2000 the same separation is attained by means of a prism only 4 cm. in thickness. CHAPTER III. SPECTROSCOPES. THE numerous forms of instruments for spectrum analysis are all divisible into two classes, prism spectroscopes ' with angular or direct vision, and grating spectroscopes. 3 The forms vary according to the special purpose for which the instrument is to be employed, such as exact measurements, quantitative and photometrical investigations, microscopical or astronomical observations, or for the preparation of spec- troscopic photographs. Prism Spectroscope with Angular Vision. The appa- ratus employed by Kirchhoff and Bunsen 3 in their earlier investigations is shown in Fig. 4. It consists of a hollow glass prism F filled with carbon bisulphide, of a telescope C magnifying eight times, and of the slit tube or collimator B, at the end nearest to the light; this is closed with a plate pierced with a fine slit; the other end contains a lens which makes the light-rays coming from D parallel before they fall on the prism. Shortly afterwards Steinheil of Munich con- 1 For information on the theory of the prism in the spectroscope see Reusch, P. A. (1862) 117, 241. Pickering, Sillim. Journ. (1868) 45, 301. Christie, P. R. S. (1877) 26, 9. Thollon, Journ. de phys. d'Almeida (1878), 7, 141. 2 For the theory of gratings see Rowland, P. M. (1892) [5] 13; 16, 197. Astronomy and Astrophysics (1893), 12, 129. Ames, Johns Hopkins Univ. Circular (1889), 8, No. 73, p. 69. P. M. (1889) [5] 27. Runge, Winkel- mann's Handb. d. Phys. (Breslau, 1894), p. 407. Glazebrook, P. M. (1889) [5] 27. Mascart, Journ. d. phys. d'Almeida (1883) [2] 2. Lord Rayleigh, P. M. (1874) [4] 47. 3 Chem. Analyse durch Spectralbeobachtungen, P. A. 110, 167. SPECTROSCOPES. 21 structed for them an improved form of spectroscope (Fig. 5) which is still in use. A flint-glass prism P of 60 is fastened to a cast-iron stand which also carries the collimator-tube A, FIG. 4 . FIG. 5. the telescope B, and the telescope C containing a scale. The mechanism for producing the slit carried by the collimator- tube is shown enlarged in Fig. 6. The width of the slit can 22 SPECTRUM ANALYSIS. be regulated by a micrometer-screw; on the lower end a small reflecting prism (Fig. 7) is fixed, by means of which a second source of light, placed at the side, may be examined together with the first. With the apparatus arranged as in Fig. '5 the spectrum of the flame F appears above that of/, so that it is possible at a glance to see whether the former contains the substance sought, if a specimen of it is simultaneously volatilized in the latter. A millimetre scale 5 is contained in the tube C\ it is illuminated by a small luminous flame, and FIG. 6. FIG. 7. its image reflected by the adjacent side of the prism into the telescope B. In order to prepare such a spectroscope for use, the telescope B is detached, and adjusted to infinity by observing some fixed object at a considerable distance; if it contains cross-wires, these must be first focussed with the eye- piece; the telescope is then replaced, and the slit opened so that the spectrum lines are sharply defined, until, for instance, the sodium lines are resolved, and the scale-tube is drawn out to make the divisions clearly visible. In many modern instru- ments the length of the tubes A and C is adjusted to their lenses before they are sent out, so that only the eyepiece requires focussing by each individual observer. The clearness of a spectrum is considerably influenced by the slit apparatus. The manufacture of these has greatly im- proved in the course of time, in finish, in the opening mechanism, and in the permanency of the material employed; the best substance for the edges is quartz, but platinum or brass is generally used. The spectroscope above described is suitable for chemical laboratories, but not for astronomical purposes, for which a greater dispersion is necessary; this may be obtained by the use of several prisms instead of one. The instrument made by Steinheil, and used by Kirchhoff in preparing his draw- ings of the solar spectrum, is shown in Fig. 8. It contains FIG. four prisms, and is at once simple and sensitive. The prisms are of flint glass with angles of 60, 45, 45, 45; the light traverses them successively, and is refracted through 130, so that the spectrum is greatly extended. The telescope B enlarges 36 and 72 times, according to the lenses employed, and moves on a divided circle by means of a micrometer-screw R\ with the help of the cross-wires the distance between two lines may be readily measured; the slit is regulated by a sliding micrometer, and is provided with a comparison-prism. The measurements may also be made with an illuminated scale in a second telescope (not shown in the figure); this is sometimes inconvenient, as the spectrum is so much longer 2 4 SPECTRUM ANALYSIS. than the scale that the latter requires frequent readjustment. In the above instrument the prisms are moved by hand into the position of minimum dispersion for any given color; this is unsatisfactory: the arrangement devised by Browning (Fig. 9) makes the adjustment automatic. The prisms are FIG. o. connected to each other, and to the observation-telescope, by hinges, and are fixed on metal plates the other ends of which are suitably cut to receive a central screw ; on rotating the telescope B to any particular line in the spectrum, the prisms move with it, so that the ray traverses then symmetrically. Gassiot's spectroscope, constructed by Browning on this plan, contained nine prisms and possessed high refractive power. There are several other ways by which the dispersion may be increased; instead of using several prisms the light may be repeatedly passed through a single one, or hollow prisms filled with some liquid of high refractive power may be employed. SPECTROSCOPES. Acting on a suggestion of Littrow, 1 Grubb, 1 C. A. Young, and Lockyer caused the light to pass twice through the same FIG. 10. prism, as shown in Fig. 10. The ray passes first through the upper part of the prism and is then returned through the lower portipn by means of a reflecting prism. The instru- ment shown is made by Browning; it is extremely powerful, and suitable both for laboratory and stellar work. It contains six prisms, besides the reflecting prism, and consequently the dispersive power is equal to that of twelve prisms. The posi- tion of any one prism can be altered at will, without interfer- ence with the other parts of the instrument, so that the dispersive power can be readily changed from two up to twelve prisms as required. The adjustment of the prisms to the position of minimum dispersion is made automatically in 1 Wien. Her 47, 2, p. 26. 8 Monthly notices of the Roy. Astron. Soc. 30, 36. 26 SPEC TR UM A NA L YSIS. the manner described above; the position of the spectrum- lines is measured by means of a micrometer-screw, the move- ment of which also adjusts the prisms. Hilger of London constructed a large spectroscope in which the ray of light was passed six times through the same prism. The application of this method, and also the number of prisms which can be employed, is limited; the longer the spectrum the less bright any given portion must necessarily be. Moreover, the glass of which the prisms are composed is never absolutely homo- geneous, nor the faces perfectly true; hence when many prisms are used there is a loss of clearness and definition. Ordinary prisms are greatly surpassed in dispersive power by the compound prism of Browning and of Rutherfurd, 1 which bears the name of the latter. It consists of a flint-glass prism FIG with its faces at such an angle that light which enters cannot emerge; in order to permit this, compensating prisms of crown glass, and therefore with a lower dispersive power, are cemented to each side face. This use of crown glass has (Comparatively little effect on the dispersion. vThollon 2 constructed an instrument which, whilst probably 1 Sillim. Journ. (1865) [3] 35, 71, 407. 2 C. r. 36, 329, 395, 595; 88, So; 89, 749. SPECTROSCOPES. 2J unsurpassed in dispersive power, contains only a small number of prisms (Fig. 1 1). Its efficiency is due partly to careful calculation of the most suitable angles for the faces of the prisms, partly to their being filled with carbon bisulphide, which has a high refractive power (comp. preceding chapter). Only compound prisms are used, one of which is shown in Fig. 12. The refractive angle of the inner flint- glass prism is 90, that of the carbon bisulphide prism 113, and of the FIG. 12. crown-glass prism 18 and 31 respectively. The light passes from the collimator CBA through the compound prism A (Fig. 13), then through the half-prism B, to the reflecting FIG. 13. prism P\ it now returns at a lower level, traversing the sym- metrically arranged system A 'B'P', and finally emerges below the prism A. The telescope (E, Fig. u) of this instrument is fixed, the prisms being movable, and maintaining the posi- tion of minimum dispersion. The screw F serves to rotate the prisms, and also a strip of paper, such as is used in the Morse telegraph-instruments; when a line is observed at the point of intersection of the cross-wires the lever D is pressed, causing the marker to record the position on the paper. Prisms containing carbon bisulphide are all subject to the disadvantage that its refractive power is greatly influenced by temperature, an increase of o. i C. being sufficient to alter the position of the lines to a distance equal to that between 28 SPECTRUM ANALYSIS. the two sodium ZMines. It is therefore necessary, when using such prisms, to take care that the temperature remains uniformly constant. To accomplish this, Rayleigh, and also Draper, employ automatic stirrers. It has been proposed to replace the carbon bisulphide by other highly refractive liquids; Wernicke has suggested ethylic cinnamate, and Walter or-bromo-naphthalene as being suitable for this pur- pose. Direct-vision Spectroscopes. With the spectroscope last described the emergent ray travels at an angle with the entering one; instruments in which the slit, lens, prism, and telescope are in a straight line are termed direct-vision spec- troscopes. They can be readily attached to a microscope or telescope, they are easy to handle and transport, are compara- tively cheap, and the source of light can be viewed directly, so that these advantages over the ordinary form have led to their wide use for practical purposes where a great dispersion is not required. Amici in 1860 constructed a compound prism which almost permitted of direct vision; it consisted of a flint-glass prism of 90, with a crown-glass prism on each side; Janssen 1 afterwards made a more elaborate system (Fig. FIG. 14), consisting of three crown- and two flint-glass prisms, which, whilst resolving the ray into "its constituents, prevents its deflection. The spectrum from a flint-glass prism is almost double the length of that from a similar one of crown 1 C. r. 55, 576. SPE C TROS COPES. 2 9 glass, so that the dispersion is reduced by about one half in consequence of refraction in the opposite direction. It is only in a certain definite part of the spectrum, generally the green, that the incident and emergent rays follow the same path; the other portions are deflected to each side. The spectroscope constructed by Hofmann of Paris, under the FIG. 15. direction of Janssen, is shown in Fig. 15. The slit S of steel, is regulated by means of a screw, and has a comparison- prism; the tube P contains the lens E and a compound prism (Fig. 14); it is attached to the telescope F, which can be adjusted to any portion of the spectrum by the screw X. For most practical purposes a telescope is unnecessary. The instrument (Fig. 16) first constructed by Browning of London is very convenient, and is widely used; it is known as the direct-vision, pocket, or miniature spectroscope. It consists SPECTRUM ANALYSIS. of the slit S, the size of which can be regulated by turning the cap s, the lens C, and the compound prism P, composed of four crown- and three flint-glass prisms; the eyepiece R FIG. 16. is adjustable. Another form of the same instrument (Fig. 17), also made by Browning, carries a detachable comparison- FIG. 17. prism and a photographed micrometer-scale, which, together with a biconvex lens, is contained in a small tube fixed parallel with the larger tubs by a slot attachment; a reflection- prism throws the image of the scale on the outer surface of the last member of the compound prism, whence it is reflected into the eye of the observer. The instrument is only 8.5 cm. in length, and when in use is attached to a readily adjustable stand. Pocket spectroscopes similar to the simpler Browning form are manufactured by all instrument-makers. Guided by H. W. Vogel, 1 Schmidt and Haensch of Berlin construct an instrument which, instead of the scale, has a small rotatable mirror; the light from this is projected on to a reflecting prism, and thence to the upper portion of the slit. The mirror and comparison prism can be readily disconnected 1 Ber. 9, 1645; 10, 1428. SPECTROSCOPES. 31 The majority of the above spectroscopes give a field extending only from the A- to the (9-line; the violet portion is almost completely absent. Adam Hilger of London con- structs a direct-vision spectroscope which, whilst somewhat longer than the others, is characterized by a high dispersive power. The spectrum extends from the extreme red beyond the //-line, shows the two /Mines, and, when directed at the sun, the nickel lines between them. The instrument is fitted with an achromatic eyepiece, and a special arrangement which reflects a slender line of light on to the spectrum to serve as a means of measurement; as the color and intensity of the line can be regulated, it permits, particularly in the darker portions of the spectrum, of far more accurate determinations than could be made with cross-wires. The measuring is per- formed with the help of a micrometer-screw attached to the slit, which it moves from end to end of the spectrum. Amongst other direct-vision spectroscopes l Christies' 2 deserves mention on account of peculiarities in its construc- tion. He employed "half-prisms," which are so called because they may be regarded as the halves of a tripartient compensating prism; the refracting angle may vary: if it is 90, a long enlarged direct-vision spectrum is obtained, but the dispersion is not correspondingly great, and the illumina- tion is poor. Hitherto it has only been used in England. Grating Spectroscopes. The advantages of diffraction- spectra have been already numerated, and it was mentioned that the grating spectroscope can only be used where the light is extremely bright, as in the case of the sun and elec- tric arc, since the loss of light is considerable. The instru- ment is usually employed in the form of a spectrometer, of which one variety is shown in Fig. 20. The gratings are of glass or metal, the latter being preferable for exact 1 Alex. Herschel, Monit. scientif. 7, 259. Emsmann, P. A. 150, 636. Kessler, P. A. 151, 507. Fuchs, Zeitschr. f. Instrumentenkunde, 1, 352. 2 P. R. S. 26. 8. SPE C TR UM A NA L YSIS. work. The general plan of a spectrometer is shown in Fig. 18. The grating m is at right angles with the collimator L, and secured to the bed of the instrument; the telescope F is fitted with cross-wires, and situated at right angles with the axis of the instrument. The telescope is directed towards the slit d, and its position read off on the divided circle; homo- geneous light, such as the so- dium flame, is then allowed to fall on the slit, and the telescope rotated until the spectra of the first, second, and third order are successively brought into the field of view, each position being noted; the wave-length of sodium light being accurately known, the readings provide a means of measuring the wave- length of any other kind of light, since the wave-length of the latter bears the same pro- portion to that of sodium light as the corresponding scale read- ings. The wave-length can also be directly determined if the grating-constant is known ; this is effected by counting under the microscope the number of lines in I mm. It was o with such an instrument that Angstrom made his celebrated determinations of the lines in the solar spectrum. A simple form of spectroscope with a reflex grating is shown in Fig. 19. The telescope and collimator are situated close together on separate mountings; the grating is enclosed in a case with a plane parallel glass front to protect it from FIG. 18. SPECTROSCOPES. 33 corrosive fumes, and is fixed on a revolving stand so that spectra of any desired order can be brought before the cross- wires of the telescope. The spectra of higher order overlap,, but they may be easily separated by following Fraunhofer's suggestion, and placing a prism between the grating and tele- FIG. 19. scope in such a position that the plane of refraction is at right angles to that of the grating; the spectra then appear clearly one above another, and can be separately observed. Rowland's Concave-grating Spectroscope. The de- velopment of spectrum analysis received a considerable impetus, from Rowland's discovery of the concave grating in 1881. By its use measurements have attained a degree of accuracy otherwise unapproachable, and whilst this is specially true of the values obtained by the coincidence method, it also applies, to wave-lengths directly determined ; it is the only instru- ment which is available for use with all rays, including the ultra-violet and the infra-red, and, as no lens is necessary between the slit and eyepiece, defects from loss of light or spherical aberration are avoided ; the gratings being astig- 34 SPECTRUM ANALYSIS. matic a luminous point, such as a spark, appears in the field of view as a line, thus greatly facilitating the comparison of solar lines with those of metals, and the enlargement of spectra. Photographs, both of the visible and invisible por- tions of the spectrum, are easily obtained, and their accuracy is necessarily far in excess of the drawings prepared from ocular observation ; as it is generally used in conjunction with a camera, its detailed description is reserved for the following chapter. When employed for ocular purposes, the camera is replaced by a cross-wires and micrometer; with a highly accurate screw of 125 mm. the measuring arrangement resem- bles a dividing-machine in exactitude rather than an ordinary micrometer. The use of an eyepiece with a focal length of -J inch gives results equalling those obtained with a plane grating in combination with a telescope enlarging 100-200 times. CHAPTER IV. SPECTROSCOPIC INSTRUMENTS FOR SPECIAL PURPOSES. Spectrometer. This instrument is employed when exact measurements are required in spectrum analysis, the coeffi- cient of refraction being determined by the method of mini- mum dispersion. The numerous instruments on the market, whilst differing in detail, agree in principle; the one described here (Fig. 20) is constructed by Schmidt and Haensch of Berlin, according to the design of V. v. Lang. The stand can be adjusted horizontally by means of levelling-screws; it carries in the middle a pillar with a steel axis; in the centre at the upper end of this is a metal plate with a divided circle, at the lower end a clamp and six radial arms for the rotation of the axis; fitting over the central pillar is a stout bronze cylinder which carries the vernier circle at the top ; an astronomical telescope with cross-wires and counterpoise is fastened to the side, whilst the clamping arrangement is attached below. The stand carries the micrometer-screws for the divided circle and vernier, and also the holder for the collimator, which is fitted with a lens, an adjustable slit, and a comparison-prism. Rising from the central pillar through the middle of the divided circle is a rod carrying a round plate, the height of which can be regulated; it also is divided, has a fixed vernier, and serves to support the prism or grating. The circle is enclosed in a case fitted with windows, so as to protect it from corrosion; both it and the outer vernier-circle can be rotated independently; to facilitate readings the tele- 35 SPECTRUM ANALYSIS, WMMJ l^-*!>-^ ..*-' I INSTRUMENTS FOR SPECIAL PURPOSES. 37 scope may also be moved without altering the position of the circle. The instrument is provided with a Gauss' eyepiece (Fig. 21), in addition to one of the ordinary form ; it has an opening (b) at the side, which admits light to a plane parallel glass plate placed behind it at an FlG * 21> angle of 45 with the axis of the telescope; the light is thus reflected on to the cross-wires. The correct orientation of the instrument is of considerable importance. The telescope is adjusted to infinity and placed at right angles to the axis of rotation, the cross- wires and eyepiece being in sharp focus; the collimator must also be adjusted to infinity and fixed at right angles to the axis of rotation, whilst the refracting edge of the prism is parallel with this. The refractive index () is calculated from the angle of refraction (g) and that of dispersion (D). In order to obtain the former, the telescope, fitted with the Gauss eyepiece, is adjusted at right angles to the first prism-face, and then rotated until it. is at right angles with the second face; the reading obtained, subtracted from 180, gives the refractive angle (g). The angle of dispersion (D) is determined as follows: The prism is removed, and the telescope with cross- wires directed on the slit this gives the zero-point ; the prism is then replaced in the position of minimum dispersion, and the telescope again adjusted towards the slit the difference in the readings gives the angle of dispersion for the particular color. The coefficient of refraction is then calculated by the expression sin sm and the wave-length of the refractive index n by Cauchy's * dispersion-formula B C n = A + - + + . . . ; Memoire sur la dispersion de la lumiere (Prague, 1836). 30 SPECTRUM ANALYSIS. for ordinary purposes the last constants can be neglected, as they are very small. A and B are obtained from the expres- sions B . B V A 2 " which necessitate the determination of the refractive indices n l and # a of the prism for two rays, the wave-lengths A t and A 2 of which are accurately known, such as two prominent Fraun- hofer lines. Kruss' ' Universal Spectroscope. This instrument is manufactured by A. Kruss of Hamburg from the design of G. Kruss; 1 it is suitable for all kinds of spectro-chemi- cal investigation, and admits of measurements being made which in accuracy closely approach those obtained by the spectrometer. In its general plan the instrument (Fig. 22) FIG. 22. resembles that of Bunsen and Kirchhoff; the telescope mag- nifies seven times; the tube B carries the scale, which is fixed in the focus of the objective; the 100 division is adjusted to correspond with the middle of the ZMines; the slit on the collimator A is likewise sharply in the focus of the objective> 1 Ber. 19, 2739. INSTRUMENTS FOR SPECIAL PURPOSES. 39 and parallel with the refracting edge of the prism. Two different slits are employed, a single one for qualitative, and a double one for quantitative analysis; in the case of the latter both slits open symmetrically to the optical axis, so that with apertures of all sizes the spectra retain the medial position. The single slit is provided with, a detachable comparison prism, and its width regulated by a micrometer screw with a divided milled head. The halves 5 : and 5 2 of the double slit 5 may be separately adjusted by means of micrometer-screws with the divided milled heads t l and A,. Two prisms are pro- vided, one of flint glass of low dispersive power with an angle of 60, and a highly refractive Rutherf urd prism ; they are contained in the case D, and are retained in the position of minimum dispersion by the pressure of a spring under the knob K. The measuring appliances are attached to the tele- scope, which, together with its holder, may be rotated around the vertical axis of the instrument by a micrometer screw with a milled head r^ divided into 100 parts; complete revo- lutions of this are read off on a divided scale directly under the eyepiece. The cross-wires are moved independently by a similar screw with its milled head r 2 also divided into 100 parts. These arrangements permit of accurate measurements of spectra; and as the relationship of the screw-threads to each other and to the divisions of the scale are known, the measurements made by any one can be doubly controlled. Spectrophotometer. A special peculiarity of the pre- ceding instrument is the double slit, first employed by Vierordt l in his photometric work with absorption-spectra and quantitative spectrum analysis. To determine the quantity of a colored substance in solution it is poured into a trough or a Schultz's glass, and placed in front of one slit; the other is then closed until both are of equal brightness; the move- J P. A. 140, 172. Die Anwendung des Spectralapparates zur Photo- metric der Absorptionsspectra und zur quantitativen Analyse (Tubingen, 1873). 40 SPECTRUM ANALYSIS. ment of the slit is measured by the divided screw-head, the amount of light cut off being equal to that absorbed by the quantity of substance in solution. In Glan's, 1 and Htifner's 8 spectrophotometers the diminution of the light is not deter- mined by closing the slit, but by polarization. Sorby's Microspectroscope. It is often desirable to apply the spectroscope to the investigation of microscopic objects, such as rock sections, leaves, the sap of plants, opaque substances, and for the identification of the coloring matter of blood in medico-legal cases. For this purpose Sorby, 3 in conjunction with Browning, arranged a convenient combination of microscope and spectroscope. The instrument (Fig. 23) is inserted into the tube of a microscope instead of the ordinary eyepiece; the rays proceeding from the lens pass successively through the slit, the combining lens, and an Amici compound prism; at right angles is placed a measuring arrangement consisting of a ray of light from a mirror which is reflected into the eye of the observer by the exterior face of the last prism ; a dark photographed background is pro- vided, and a delicate micrometer-screw which enables the ray to be accurately adjusted to any spectrum or absorption line. A second mirror serves to illuminate the comparison-prism, and a small stand at the side on which objects can be fixed. Light for the mirrors is obtained from a single lamp. A modification of this microspectroscope devised by Abbe and constructed by C. Zeiss of Jena is shown in Fig. 24. The screw M fixes it in the tube of the microscope in such a position that both the mirrors A and O are illuminated by the sun. The upper portion, containing the prisms, can be rotated round the peg K so as to admit of the object being adjusted ; when this is accomplished the prisms are turned until the closing of the catch L indicates that they are in position. 1 W. A., i, 351. 9 J. pr. Chem. [2], 16, 290. Zeitschr. f. phys. Chem., 3, 562. 3 C. N.. 15, 220. P. R. S., 15, 433. INSTRUMENTS FOR SPECIAL PURPOSES. 4! The mirror^4, indicated by dotted lines, illuminates objects fastened to a stand at the side ; the light passes through an opening to the comparison-prism, which is fastened to the slit FIG. FIG. and is not shown in the figure. A peculiarity of the instru- ment is found in the measuring arrangement: the scale N gives the wave-lengths, in parts of a micromillimetre, of the region of the spectrum, coincident with its divisions; the compound prism is fixed between pieces of cork, and can be inclined from the vertical by means of the spring Q and the regulating-screw P, so that the divisions of the scale may be made to correspond exactly with the Fraunhofer lines; the mirror O reflects the scale on to the exterior prism-face. Solar and Stellar Spectroscopes. The extensive appli- cation of spectrum analysis to astronomical purposes has resulted in the designing of a very large number of instruments suitable for such observations. Solar investigations are gen- erally carried out with fixed apparatus, grating spectroscopes, 42 SPECTRUM ANALYSIS. spectrometers, and angular-vision spectroscopes of high dispersive power, the light being supplied by means of a heliostat. The solar prominences could formerly only be observed during an eclipse, but Lockyer and Janssen simul- taneously devised a method by which they may be investi- gated at any time. For this purpose a spectroscope of high dispersive power is required; it is connected with a telescope, and the slit of wide aperture directed tangentially to the solar limb. Any spectroscope of high dispersive power may be employed for this purpose, if it admits of rotation around the collimator-axis so as to discover the prominences. It is not easy to securely fasten a heavy spectroscope to a telescope ; for this purpose a special form of stand, termed an adapter, has been described by H. C. Vogel, but it is highly desirable that the spectroscope should be as light as possible. Such a FIG. 25. solar spectroscope, designed by Browning, is shown in Fig. 25 ; an arrangement is provided by which it can be screwed on to the eyepiece of a refractor of three or more inches diameter. The screw-ring carries a position-circle, divided into whole degrees, the pointer (alidade) of which has two adjustable bars at right angles; to the front one are attached the sup- ports for the collimator and prism-plate. The collimator and telescope are fixed, the five prisms move automatically; a INSTRUMENTS FOR SPECIAL PURPOSES. 43 reflection-prism causes the rays to return through the system, so that the effect of ten prisms is obtained. The measuring arrangement is on the left of the telescope; it consists of a triangular brass rod with a micrometer-screw by means of which the prisms are automatically adjusted to the position of minimum dispersion for the particular ray under examina- tion. The instrument shows six lines between the two ZMines with the sun at its brightest. Since the dispersive power can be varied from two to ten prisms, the instrument can also be used for stellar and other celestial observations, where, in consequence of the feebleness of the light, a lower dispersion is required. For a general survey of the sky, or for the observation of fixed stars which appear as points, a telescope may be at- tached to a direct-vision spectroscope without a slit; such instruments have been designed by Zollner, Secchi, H. C. Vogel, McClean, v. Konkoly, and others, but they are not adapted for accurate measurement. For the investigation of the spectra of meteors, v. Konkoly recommends a direct-vision spectroscope with a concave cylindrical lens attached to a small telescope; the field of view then includes about 27, and the rapid motion of the meteorite is apparently diminished. Stellar Spectrometers. H. C. Vogel has designed a good spectrometer which is fastened to a telescope, and permits of accurate stellar spectroscopic measurements being made. More recently Adam Hilger of London has con- structed an instrument specially adapted for solar and stellar work, and it can also be mounted on a stand for use in the laboratory. Spectrographs. The use of a combination of spectro- scope and photographic camera, termed a spectrograph, has led to extremely valuable results: errors of observation and drawing are avoided; records are obtained of the invisible infra-red and ultra-violet regions, and also of celestial objects, 44 SPECTRUM ANALYSIS. the light from which is too feeble to affect the eye, and necessitates a prolonged exposure. H. W. Vogel 1 has designed a large and a small form of this instrument; the latter has a wedge-shaped slit, a collimator-lens, an Amici prism of five members, and an aplanatic lens which throws the image on the ground-glass or sensitized plate at the back of the camera. The larger instrument is adapted for deflected rays, has two prisms of 60, and is rotatable around both the vertical and horizontal axes, which permits it to be used, for solar and terrestrial purposes. Ostwald a has suggested the following procedure for the photography of absorption-spectra : A horizontal plate is screwed on to the objective of a photo- graphic camera, to which a spectrometer-prism and collimator- tube containing the lens are attached ; the spectrum is thrown on the ground-glass plate by a Luter's aplanatic lens of 40 cm. focal length, and brought to accurate focus by obser- vation of the solar lines. The prism employed has a refracting angle of 60, and is filled with ^-bromonaphthalene; 3 its dis- persive power is especially high in the ultra-violet. An Auer incandescent gas-lamp is recommended as the source of light. Photographs of stellar spectra are obtained by attaching the spectrograph directly to a telescope. The large spectro- graph at the Astrophysical Institute at Potsdam 4 is an exam- ple of such an instrument; it gives extremely accurate reproductions of stellar spectra. The ocular of an eleven-inch refractor is removed, and a strong stand attached in its place; to one end the spectrograph is screwed; the cbllimator-tube of this, in order to insure stability, is sustained on a conical steel holder with an adjusting arrangement and a scale. Next to the collimator is a stout circular holder for the two Ruther- fiird prisms of high dispersive power; beyond this is a conical 1 P. A. 154, 306; 156, 319. 2 Zeitschr. f. phys. Chem. 9, 579. 3 Walter, W. A. 42, 511. 4 Scheiner, Spectralanal. der Gestirne (Leipzig, 1890), p. 109. INSTRUMENTS FOR SPECIAL PURPOSES. 45 camera which, to guard against vibration, is attached by stays to the far end of the collimator-tube. The collimator and object lenses are achromatic to chemically active rays. About 40 cm. from the slit, in the focus of the refractor objective, is placed a Geissler's tube containing hydrogen ; this serves to illuminate the slit, and also affords a means of measuring the displacement of the lines caused by the motion of the star across the field of view, since the photograph shows the stellar spectrum together with the //^-lines. The refractor is set in motion so as to maintain the middle of the slit exactly on the star. Bromo-silver gelatine plates are usually employed. Rowland's Concave-grating Spectrograph. 1 The merits of this instrument have been enumerated in the preceding FIG. 26. chapter; its separate parts require special arrangement, for, as Rowland has pointed out in his theory of the concave grating, normal spectra are only obtained if the slit, grating, and camera, or eyepiece, are placed in the periphery of a 1 P. M. [5] 16, 197. Astronomy and Astrophysics (1893), 12, 129. Johns Hopkins University Circulars (1889), 8, No. 73, P- 73- Ames, ibid., 8, No. 73, p. 69. P. M. (1889) [5] 27. 4 6 SPECTRUM ANALYSIS. circle of which the grating forms a segment; the diameter of the circle is dependent on the concavity of the grating. This adjustment is effected mechanically by means of two rods at right angles; the grating is placed at their point of intersec- tion, so that it is always maintained at a constant distance from the camera. The external appearance of the instrument is shown in Fig. 26. Fig. 27 exhibits the plan of the one FIG. 27. employed by Rowland for his classical measurements of the solar lines. It consists of two stout wooden beams AB and 15 X 33 cm. and 7 metres in length; one is fixed, but INSTRUMENTS FOR SPECIAL PURPOSES. 47 the other is slightly movable, by means of screws, about the point A. The beams are provided with rails for the wheels at each end of the carriages D and E \ apertures are cut in these ex- actly coincident with the rail to receive the peg of the trans- verse beam FG, which is made of 4-inch wrought-iron tube; its length corresponds with the concavity of the grating, and, in the case of a six-inch grating with 1 10,000 lines, is about 6.55 metres, with 15 cm. for adjustment. The grating is fixed FIG. 28. to D, and the camera to E\ the latter is shown in Fig. 28. It consists of the fixed box B, and the movable case A to hold the photographic plate, which is bent to the desired form by pressure against a rubber pad; the wooden rod C carries a brass plate with a long narrow opening; it is rotatable about its horizontal axis, and is used to obtain a comparison-photo- graph. The sensitive plates are 48 X 5 cm., and 1.8 mm. in thickness; the breadth of the spectrum varies according to the order, from 6 mm. to 10 cm. The grating-holder is shown in Fig. 29, and consists of a massive bed-plated, and a mov- able perpendicular frame B\ to this a brass rod D is fastened by the screw P, and is movable about this axis by the SPECTRUM ANALYSIS. screw 5; a.second movable frame is attached to the brass rod by means of the screw P, and is rotatable about P' by the ) p p b 1 3 ' FIG. 29 FIG. 30. screw S'; it carries the grating, supported on two projec- tions, and fixed by wax so as to be quite free from pressure, The orientating-screws work against springs so that the grat- ing is readily movable in any direction. Gratings of varying diameter are used with 10,000 to 20,000 lines to the inch; for most purposes the first is sufficient, but the last is preferable for work with the ultra-violet rays. The slit (Fig. 30) must be capable of being raised, lowered, and rotated about its horizontal axis; the latter is of special importance, as a sharp image can only be obtained if the slit and lines are exactly parallel. The aperture is regulated by means of a micrometer- screw; as a*rule its width is only 0.025 mm - The light from the electric spark or arc, or from a heliostat if the solar spec- trum is being examined, is directed on a convergent quartz lens in the focus of the slit, whence it passes directly to the slit; if the source of light is at the side, a reflecting prism is employed. The astigmatism of the grating precludes the simultaneous photographing of a spectrum and a superposed comparison-spectrum, as with the prism-spectroscope; the comparison-photographs are therefore made separately, but on the same plate. By means of the arrangement at the back INSTRUMENTS FOR SPECIAL PURPOSES. 49 s of the camera, the comparison-spectrum is received in the middle of the plate, and the spectrum under examination; above and below. A triangular stand placed between the slit and the quartz lens serves to support vessels containing absorbent solutions to cut off the spectra of other orders which would otherwise obscure the one employed. These concave-grating spectroscopes, of excellent finish,, are manufactured by John A. Brashear of Allegheny. By means of Rowland's coincidence method a knowledge of one absolute wave-length enables ah others to be deter- mined, relatively to this, with an extremely high degree of accuracy. 1 Taking D l as the basis, it is photographed in the spectra of as many orders as the grating permits; the other orders are thus obtained simultaneously on the same plate, viz., D l in the first order, 2948 in the second, and 1965 in the third; with D l in the second order, 3931 in the third, 2948 in the fourth, 2358 in the fifth, etc. The following table gives the coincidences with > l in the first nine orders: I 5896 2 2948 5896 3 I9 6 5 393i 5896 4 2948 4422 5896 5 2358 3538 4717 5896 6 1965 2948 3931 49T3 5896 7 2527 3369 4211 5054 8 2211 2948 3685 4422 9 2620 3275 3931 i The distance between the Z>,-line and separate lines of these different orders is measured, and the wave-lengths approximately calculated from the grating constant, which is known with i fair degree of accuracy, and its concavity; if the wave-length of the lines does not differ by more than 50 1 Comp. Kayser, A. B. A. 1890. 50 SPECTRUM ANALYSIS. Angstroms from that of Z>,, a result is immediately obtained, correct within 0.5 of an Angstrom. The accuracy of this result may be increased to any desired degree; if, for exam- ple, the wave-lengths 4422 and 47 17 0.5 are photographed on a plate, and their distance measured, the result is accurate to one third per cent, since their difference of 300 units is known to be exact to one unit; with this higher approxima- tion the calculation is repeated, giving data of still greater accuracy, and so on. Rowland obtained his system of wave-lengths in this manner; between A 2400 and A 7000 it is exact within at least o o.oi Angstrom, and, together with his atlas of the solar spectrum, it forms the basis of all spectroscopic measure- ments. CHAPTER V. SPECTROSCOPIC ADJUNCTS. Flame-spectra. The flame of a Bunsen burner is suffi- cient to volatilize compounds of the metals of the alkalis and alkaline earths; for compounds of these metals which are par- ticularly difficult to vaporize a Terquem burner may be used, as recommended by Wiedemann and Ebert. The flame, spectra of many substances can only be obtained by the use of a blowpipe, hydrogen, or oxyhydrogen flame, all of which were employed by Kirchhoff and Bunsen. A Barthel or other suitable spirit-lamp may take the place of a gas-burner if necessary. In many cases a relatively cool flame is desir- able; this can be obtained, according to Salet's suggestion, by placing the gas-flame in contact with a plate of marble or metal, the opposite side of which is kept cool by means of a 52 SPECTRUM ANALYSIS. stream of water. The substance is brought into the flame on a pbtinum wire; one end is bent to a -loop, and the other is fused into a small piece of glass tubing which is fastened to the arm of a Bunsen stand (Fig. 31). IVIitscherlich's appa- ratus is employed when a flame-spectrum is required during a considerable period of time. It consists of a series of glass tubes (Fig. 32) closed at the upper end ; the lower end is bent, and filled with a bundle of slender platinum wires or threads of asbestos c\ the substance under examination is dissolved, a little hydrochloric acid or ammonium acetate added, to pre- vent the formation of crusts on the wires, and the solution placed in the tubes. Gouy's 1 method consists in causing the air or gas, before it reaches the burner, to pass through a vessel containing the substance in solution or in a very finely divided state. Neither of these arrangements is satisfactory when a uniformly bright flame has to be maintained during 1 A. c. p. [5] 18, 25. The construction of the pulverizing apparatus is described in Ebert's work Anleitung zum Gasblasen (Leipzig, 1895, 2d ed.) P- 59- SPECTROSCOP1C ADJUNCTS. 53 several hours; for such purposes Eder and Valenta's 1 appa- ratus (Fig, 33) can be employed. It consists of a heavy metal pedestal P, rotatable about its vertical axis; the Bunsen or Terquem burner # is fitted with a platinum ring; the wheel s is adjustable, and inclined at an angle of 45: it consists of two nickel plates carrying a circle of platinum gauze n, which projects 2-3 cm. beyond the circumference of the wheel, and FIG. 33. dips into the vessel containing the solution of the substance. The axle a, connected with the cone c, permits of the rota- tion of the wheel by clockwork or some suitable motor. Mitscherlich,* Wolf and Diacon, 3 and Salet * place readily volatilizable substances in a glass tube fused to the hydrogen generator, and heat them in the gas which is then burnt at a platinum jet fused to the other end of the tube. For pro- 1 Denkschr. der mathem.-naturw. Classe der Wien. Akad. (1893) 60, 468. 2 P. A. (1863) 121. 8 Mem. de 1'Acad., Montpellier, 1863. 4 A. c. p. (1873) [4] 28. Spectroscopie (Paris, 1888). 54 SPECTRUM ANALYSIS. longed observations on haloid componnds of metals, the coal- gas must be mixed with chlorine, bromine, or iodine vapor. The Bunsen burner gives a temperature of about 2300, and none of the others are hot enough to show more than a comparatively small number of spectra; the electric arc and spark are of much wider application. The Electric Arc. The temperature of the arc ranges from 3OOO-35OO C (Violle), and is the most generally appli- cable of all methods of producing metallic spectra, as its high luminosity permits the use of gratings, and also renders it suitable for projection. Kayser and Runge employed it in their accurate measurements of the spectra of terrestrial sub- stances; usually they used a current of 50 volts and 20-30 amperes, but occasionally, for the shorter wave-lengths, 40 amperes were required. The anode carbon is placed below, as it burns to a cavity which serves to receive the substance, and its temperature is higher than the cathode. In order to prevent the metal or salt from overflowing the cup, rods of not less than 2 sq. cm. are employed. The formation of oxide may be prevented by boring the upper carbon, and conducting a current of some gas through it, or, better, by the use of a block of gas-carbon, quicklime, marble, or magnesia as suggested by Liveing and Dewar. 1 The block is pierced completely through in two horizontal directions crossing at right angles in the middle; glass tubes carrying the carbon rods are fitted to two opposite openings; light is emitted by the third, and the fourth serves for the admission of gas (carbon dioxide or hydrogen); a fifth opening, immediately above the point of intersection of the others, is for the intro- duction of the substance. The arc plays on the body under examination and volatilizes it; Kayser and Runge recommend the use of a magnet to cause the arc to impinge directly on the substance below the opening. The apparatus is also 1 P. R. S. (1879). Proc. Cambridge Phil. Soc. (1882) 4. Kayser and Runge, A. B. A. 1890, 1891. * SPECTROSCOPIC ADJUNCTS. 5S useful for observing the reversal of the lines by absorption, as the opening through which the light passes is always full of comparatively cool vapor; the intensity of the lines is less with the block than when the substance is volatilized directly from the carbon rod. A series of carbon bands are almost always observed above the metallic spectrum when metals or salts are volatilized between carbon terminals; this often adds greatly to the difficulty of measuring the lines, and, in order to avoid error, an exact knowledge of these bands is necessary. The difficulty may be obviated by using terminals of the par- ticular metal under observation, but this has other drawbacks: the lamp is no longer automatically regulated, as the glowing rods of metal fuse immediately they come into contact, whilst, if the arc is extinguished for a moment, the terminals become coated with oxide, which is a bad conductor, and requires to be scraped off. The Electric Spark. The spark from an influence machine or induction-coil may also be used for the production of luminous vapor; the spectra of gases are produced solely by its help, it is more convenient than the arc, and is there- fore employed more frequently, although spark-spectra exceed those of the arc in complexity, and their nature is not at all clearly understood. The temperature of the spark varies between wide limits; E, Wiedemann 1 observed a temperature of 87,000 in a Geissler tube. The induction-spark differs with the construction of the coil; if this is composed of a large number of turns of thin wire, a high potential is obtained, whilst with shorter and thicker wire the potential is reduced, and the quantity of electricity increased. The intensity of the discharge may be heightened by the use of more powerful primary batteries, by the introduction into the circuit of one or more Leyden jars, or by increasing the length of the spark-gap. It is usual to 1 W. A. (1879) 6, 298. 5 SPECTRUM ANALYSIS. connect the poles with a condenser as shown in Fig. 34. The wire K'p' connects the outer coating of the jar L with one pole of the coil, whilst the other is joined with the spike T, whence the sparks pass to the plate P connected with the inner coating of the jar. The wires MK, M ' K r lead to the stand H, to which the metallic electrodes EE' are fastened, FIG. 34. and the spark is produced between them. The intensity and quantity of the current can also be increased when an influ- ence machine is used; the poles are connected in a similar manner with a Leyden jar, and the potential may be raised by prolonging the path of the spark. The simplest means of obtaining the line-spectrum of a -metal is to pass the spark between electrodes composed of it; particles are then detached and volatilized. The spectrum SPECTROSCOPIC ADJUNCTS. 57 of the atmosphere, consisting of the nitrogen, oxygen, and hydrogen lines, always accompanies that of the metal (comp. nitrogen, Chapter VII). In cases where a metal is not avail- able, a solution of one of its salts may be used; this is made the cathode and the spark passed over it. Delachanal and Mermet's l apparatus is very convenient for use in this con- nection, as it permits the observation of the spectra during a considerable period. It consists of a glass vessel (A, Fig. 35), 15 mm. in diameter, with a platinum wire fused through the lower end, and connected with the cathode of the coil; over it is placed a conical capillary tube D projecting 5 mm. above the end. The upper part of the vessel is closed with a cork C, through which is passed a glass tube B with a platinum wire fused to it; the end d projects and forms the anode. The salt solution is added up to half the height of the cathode ; it rises by capillarity to the end of D, and each spark volatilizes a small portion. No loss of sub- stance occurs, the slit of the spectroscope is protected from splashing, and the sparks are uniform, but they almost always attack the glass, causing the pro- duction of foreign spectra, such as those of calcium, lead, etc. Hartley's 2 apparatus (Fig. 36) is free from this drawback; the ^3* solution is poured into a U tube to one limb of which a graphite electrode is fitted, the surface having a number of deep grooves cut to facilitate the ascent of the liquid. The 1 C. r. (1875) 81. Journ. de Phys. de d'Almeida (1876), 5, 10. * P. T. (1884) 175, 49. 325- 5 SPECTRUM ANALYSIS. upper electrode may be either of metal or graphite, preferably the latter; both are chisel-shaped, connected with the coil by means of platinum wires, and placed exactly above one another and in a line with the slit. Other forms of apparatus designed for the same purpose have been described by Bunsen, 1 H. W. Vogel, 2 Lecoq de Boisbaudran, 3 Salet, 4 and Dupre/ Geissler's or Pliicker's Tubes. Plucker's method " is used for the investigation of gases; it consists in passing an electric discharge through rarefied gas contained in a Geissler tube (Fig. 37). The middle of the tube is usually a capillary; FIG. 37. platinum wires are fused through the wide closed ends, and connected inside with aluminium wires, as the passage of the current detaches particles of platinum which are gradually deposited on the glass; the light is more concentrated in the capillary, which becomes luminous, and is therefore placed in front of the slit. The tubes are rilled by means of side tubes at each of the wide portions; one is connected with a mercury- pump, and the gas introduced through the other. When the operation is completed the side tubes are sealed off; this is often attended with difficulty on account of the entrance of air due to the low pressure, 1-2 mm., which is necessary in order to secure the maximum degree of brightness. Tubes intended for observation during several hours may be closed 1 P. A. 155, 230. 2 Prakt. Spectralanalyse (Berlin, 1889). 3 Spectres lumineux (Paris, 1874). 4 Spectroscopie (Paris, 1888). 5 La nature, 1882, 220. 6 P. A. (1859) 107, 497- SPECTROSCOPIC ADJUNCTS. 59 by means of stoppers accurately ground and lubricated. 1 Cornu 2 and Deslandes 3 have given special instructions for the filling of Geissler tubes. Air is not the only impurity which is liable to be met with under extremely low pressures; the bands of carbon monoxide are frequently observed, and this is very difficult to remove. Hasselberg 1 considers that it is only occasionally derived from the fat and rubber at the joints ; it is probably liberated, together with carbon dioxide and other substances, from the glass by the electric discharge. Another disadvantage of these tubes is the thick, irregular, and non-homogeneous nature of the wall of the capillary which causes it to act as a cylindrical lens, and sometimes produces displacement of the spectrum lines. To correct this Monckhoven, 4 Piazzi-Smyth, 5 Salet, 6 Hasselberg, 7 and others have constructed tubes with the capillary at right angles to the wide parts so that it is observed "end on." In order to obviate difficulties arising from the use of elec- trodes Salet " has prepared tubes without them ; instead the wide portions are covered with tin-foil, or, if the tempera- ture is too high, with gold leaf; the tubes are connected with an induction-coil or with a Holtz machine, and the discharge takes place through influence; but under these circumstances the temperature in the tube is much lower than when the current is directly transmitted. Other difficulties attend the use of Geissler tubes: the observed phenomena may be produced by minute traces of foreign bodies, and not by the gas itself; absorption by the glass and electrodes, and the extremely high temperature 1 Hasselberg, Mem. de 1'Acad. de St. Petersb. (1883) [7] 31, No. 14. 2 J. d. phys. de d'Almeida (1886) [2] 5, 100. A. c. p. (1888) [6] 15, 28. Les Mondes, 1877. N. 19,400, 458; 20, 75. A. c. p. [4] 38, 52. Mem. de 1'Acad. de St. Petersb. (1882) [7] 30, No. 7. C. r. 73, 559. 60 SPECTRUM ANALYSIS. attainable, also influence the result, but these causes are insufficient to explain many puzzling phenomena. Liveing and Dewar, and also Ames, have suggested that these are due to the discharge producing special wave-motions of the atoms, and not the pure spectrum of the substance under examina- tion, whilst others believe that, under its influence, the molecules undergo electrolytic dissociation and the atoms are set in violent motion; it is at present impossible to give a satisfactory explanation of the matter. Observation of the Invisible Regions of the Spectrum. Only the small portion of the spectrum between wave- lengths 400^^ and 760^^ is, in ordinary circumstances, visible to the eye, but the part beyond 8oo///f becomes perceptible if the shorter waves are cut off by means of dark-red glass, whilst those far beyond 400/7^ are seen if the longer waves are eliminated. The region beyond 76o/*/* is termed the infra-red, whilst that below 4OO//// forms the ultra-violet; in the former Langley 1 reached a wave-length of 5300^, and Rubens 2 one of 5/50^/1. In the ultra-violet Schumann, 3 using gelatine plates, obtained photographs of a group of lines of A i62/^, and hydrogen lines of about ioo////. Soret 4 has devised a method of rendering the ULTRA-VIOLET visible by employing the fluorescence produced by the waves of short length; it consists in the introduction of a fluorescent object, such as a plate of uranium glass, into the eyepiece of a prism spectroscope. H. v. Helmholtz 5 accomplished the same pur- pose by placing a thin film of quinine sulphate in the telescope at the spot where the objective forms a true image of the spec- trum. Special instruments, with lenses and prisms of quartz, are required for the investigation of the ultra-violet rays, as 1 W. A. (1884) 22, 598. A. c. p. (1886) [6] 9, 433. 2 W. A. (1892) 45, 238. 3 Wien. Ber. 1892. Photogr. Rundschau, 1890, 1892. 4 Arch. sc. phys. et nat. (1877) [2] 57, 319; 61. 322; 63, 89; [3] 4, 261; 9, 513; 10, 429. A. c. p. (1877) [5] 11. C. r. (1878) 86, 1062; 97, 314, 572. 6 Optique physiologique, p. 352. SPECTROSCOPIC ADJUNCTS. 6 1 they are absorbed by glass; those of 35OyUyw to a considerable extent, and those of about SOO^yw completely. Stokes 1 recommends quartz for this kind of work, but Schumann* found that it absorbs the rays below 2OO//yu, and was obliged to substitute it by fluor-spar; his observations on waves of shortest length, referred to above, were made with such apparatus, the spectrograph being rendered vacuous. Grating spectroscopes are specially well adapted for work with the ultra-violet if the use of glass is avoided; these rays are also absorbed by the atmosphere, which accounts for the sudden extinction of the solar spectrum at 300/^5 this is extended, Cornu * found, with the sun at its zenith. Photography has latterly superseded all other methods of investigating the ultra-violet rays; those below 200/^/1 are absorbed by gelatine, and for such rays Schumann'' employed plates without a gelatine film. The INFRA-RED rays may be detected by their thermal and photochemical properties, and by means of phosphorescence. Their existence was first shown by William Herschel* in 1800. During an investigation of the heating power of various regions of the spectrum he found that the thermometer was most affected beyond the visible red. The thermal effect was shown by Melloni 6 to be influenced by the nature of the prism, rock salt being extremely readily transparent to long waves; later it was discovered that fluor-spar and sylvine are equally suitable, and the thermopile was used instead of a thermometer. 6 With a grating spectroscope all absorption by the prism is avoided, but the distribution of the heat is different, the maximum being in the yellow. In place of a 1 P. T. 1852, p 463. 2 Phot, Rundschau, 1890. 5 C. r. (1879) 88. 4 P. T. (1800) 90. 5 A. c. p. (1833) [2] 55. 6 Franz, P. A. (1855) 94. J. Miiller, ibid. (1858) 105. Lamansky, ibid. (1872) 146. Mouton, C. r. (1879) 88; 89. Desains, ibid. (1880) 90; (1882) 94. 62 SPECTRUM ANALYSIS. thermopile, Langley's ' actinic balance or bolometer is em- ployed ; by its means a rise in temperature of 0.000001 C. may be detected. It consists of a Wheatstone bridge, the arms being formed of two extremely thin blackened wires of equal resistance; if the temperature of one changes, the equilibrium is disturbed and the galvanometer affected. With the aid of this instrument Langley has examined the emission-spectra of the sun and moon, and of solid bodies between o and 1500, whilst flame, arc, and absorption spectra have been investigated by R. v. Helmholtz, 2 Julius, 3 K. Angstrom, 4 Rubens, 5 Snow, 6 Lewis and Ferry, 7 and Paschen. 8 The red and infra-red rays were for long believed to be incapable of photographic action. E. Becquerel 9 observed that the red rays affect silver chloride which has been previously exposed to light for a short time, and Draper 10 succeeded in photographing the beginning of the infra-red, but complete photographs could not be produced until Abney " prepared a special bromo-silver emulsion sensitive to the infra-red. He has obtained photographs of the solar spectrum up to wave-lengths of 2700^, both with a prism and a grat- ing, and has also photographed a number of absorption- spectra. The third method of investigating the infra-red is based ' Sillim. Journ. [3] 21, 187; [3] 27, 169; (1886) [3] 31; 32; (1888) 36; (1889) 38. Proc. Amer. Acad. 16, 342, W. A. 22, 598. Die Licht- und Warmestrahlung verbrennender Gase. (Berlin, 1890). Die Licht- und Warmestrahlung verbrannter Gase (Berlin, 1890). W. A. (1889) 36. W. A. 45, 238. W. A. (1892) 46. Johns Hopkins Univ. Circul. (1894) 13, 74. 8 W. A. (N. F.) 50, 409; (1894) 52, 209. 9 A. c. p. (1843) [3] 9- 10 P. M. (1843) [3] 22. 11 P. T. (1880) 171, 653; (1886) 177. P. R. S. (i88i).31. P.M. [5] 3, 22. SPECTROSCOPIC ADJUNCTS. / 63 on E. Becquerel's 1 discovery of their phophyrescent action. A layer of Balmain's luminous paint is exposed to diffused daylight, and then to the infra-red spectrum; at first the spectrum bands become brighter, the Fraunhofer lines re- maining unaltered; this soon changes and the Fraunhofer lines gain in luminosity until they appear bright on a dark ground. The results obtained in this manner by E. and H. Becquerel agree well with those of Langley and Abney. Lornmel 2 has improved Becquerel's method, and has pre- pared photographs of the infra-red with ordinary plates. The image of the spectrum is obtained in the manner described, with the Fraunhofer lines bluish and luminous; a gelatine dry plate is then laid over the image on the phosphorescent plate, and all the details are clearly reproduced. Photographs of the grating solar spectrum were made in this manner. The grating, particularly the concave one, in combination with a prism, is especially suitable for the investigation of the infra-red, since, by means of the superposed spectra, the wave- lengths in this region may be compared with those in the visible portion which are accurately known. The ordinary dispersion formula for prisms does not apply to the infra-red region. 3 Observation of Absorption-spectra The object under examination is placed in front of the slit; if it is somewhat opaque, direct sunlight is used ; the electric light, Linnemann's zirconium light, Auer's incandescent burner, a good petroleum lamp, or an ordinary Argand burner are also employed as sources of illunrnation. Daylight is convenient, as the Fraunhofer lines permit of ready orientation, but it is unsuit- 1 C. r. (1866) 63; (1873) 77: (1876) 83; (1887) 104; (1888) 107. A. c. p. (1877) [5] 10. H. Becquerel, C. r. (1883) 96, 97; (1884) 99; (1886) 102. A. c. p. (1883) [5] 30. - W. A. (1883) 20; (1887) 30. Sitzungsber. Munchener Akad. (1888) 18; (1890) 20. s Langley, P. M. (1884) [s] 17; (1886) 22. Sillim. Journ. [3] 27, 169. W. A. 22, 598. Desains and Currie, C. r. (1880) 90; (1882) 94. 64 SPECTRUM ANALYSIS. able for the absorption-spectra of gases, or for observations in the ultra-violet ; in these cases artificial light of great bright- ness is necessary. If the substance under examination is a gas, it is placed in a tube with plane sides. Glass troughs with plane parallel sides are used for liquids; H. W. Vogel 1 employs ordinary test-tubes, placing them so that the light- rays pass diametrically through to the slit, and, in order to overcome the difficulty of adjustment, they are fixed in a rectangular trough of water. Gladstone 2 uses a wedge-shaped vessel which allows all the absorbed parts to be observed at a glance; as a substitute Landauer 3 has suggested the use of the ordinary hollow prism, fixed horizontally or vertically in a stand; it permits of the rapid observation of various thick- nesses of liquid, and is particularly well suited for qualitative work in which the refractive angle of the vessel may be neglected. The absorption-spectrum of a substance depends on its concentration and the thickness of the column through which the light passes; this renders its accurate characteriza- tion a matter of difficulty, and is only possible when both the above factors are given. Many observers have confined themselves to giving graphic reproductions of selected charac- teristic strong and faint absorptions. Kruss 4 suggested the determination of the "minimum of brightness"; this is obtained by diluting the liquid under examination until the bands whose maxima are to be measured are readily visible; their limits are determined, and the liquid repeatedly diluted, fresh measurements being taken after each addition of water; two values are finally obtained which closely approximate if the liquid is further diluted and the minimum of brightness is situated between them. 1 Prakt. Spectralanalyse (Berlin, 1889). 2 Jour. Chem. Socy. Lond. 10, 79. P. M. [4] 24, 417. 8 Ber. (1878) 11, 1773- 4 Ibid. 16, 2051. Zeitschr. phys. Chem. 2, 312. Specielle Methoden der Analyse (Hamburg, 1894). SPECTROSCOPIC ADJUNCTS. Measuring Appliances and Scales. In the last two chapters the measuring, appliances have been described in detail together with the spectroscopes. Simple instruments with a prism of 60 have a millimeter scale reflected into the telescope; generally, following Bunsen's l suggestion, the sodium line is adjusted to the 5 E 5270.2* bi 5183-8 b, 5172.9 b 3 and b 4 . . 5168.1* F 4861.5 G 4308.0* H 3968.6 K 3933-8 Divisions of KirchhofTs Scale. 31-8 \* 50.0 /- 67.8 74-5 105-5 17-5 28.9 35-0 50.0 70.9 74-5 74-8 75-0 90.0 127.3 161.2 165.7 * Mean of several lines. SPECTROSCOPIC ADJUNCTS. 6f Photographic scales showing wave-lengths directly may be prepared, but in. certain parts the divisions are so close together that it is more convenient to use scales with equal divisions, and reduce the readings to wave-lengths. Instead of wave-lengths their reciprocals are sometimes employed, that is, the number of waves in i cm. at o in a vacuum; this value r, which is termed the oscillation frequency, is obtained by dividing into I the wave-lengths reduced to o and a vacuum, 1 the calculation being carried to six or seven figures; some observers, including Kayser and Runge, use uncorrected wave-lengths; they justified this on the ground that the refractive index of air was not known with sufficient accuracy, and that the increase in the index, as the wave- length decreases, is so small that only a negligible error is introduced by regarding it as constant; for example, the difference between A 6000 and A, = 2200 is only 0.09 Angstrom.* Large instruments, with several prisms automatically adjusted to the position of minimum dispersion, have a divided circle over which the telescope travels, thus giving the measurements, whilst the micrometer-screw, which moves the last prism, is utilized for the same purpose in the case of spectroscopes in which the rays return through the prisms, since the telescope on such instruments is necessarily fixed. None of these appliances give more than approximations, since it is difficult to obtain the mechanism absolutely exact, and the refractive and dispersive power of the prisms are affected by the temperature. In order to make accurate measure- ments with prism instruments it is desirable to compare them with the solar spectrum, or with that of iron; both are rich in 1 B. A. R. 1878. 8 Kayser and Runge, A. B. A. 1893. Hasselberg, Ofvers. K. Vetensk, Akad. Ftfrhandl. (1892) No. 9. 68 SPECTRUM ANALYSIS. lines the wave-lengths of which have been accurately deter- mined. Eder 1 recommends the use of the spark-spectrum of an alloy consisting of equal parts of cadmium, zinc, and lead for the orientation of prism-spectra of medium dispersion. The most accurate ocular observations are obtained with a spec- trometer fitted with a plane grating or prism, the best pho- tographic ones with a concave grating spectroscope. Drawings of Spectra. The diagrammatic reproduction of spectra are made, according to Bunsen's suggestion, on paper with printed millimetre scales; the breadth of the lines and bands is given as observed in the spectroscope, and their relative brightness shown by the varying distance of the contour from the horizontal; such drawings are shown in Fig. 38," together with the spectra as they appear in the spectroscope. 1 Denkschr. der Wiener Akademie (1893), 60. 2 From Wiedemann-Ebert, Physikal-Practicum (Braunschweig, 1893). CHAPTER VI. SPECTRA. I. EMISSION-SPECTRA. THREE varieties of spctra are recognized, continuous, channelled or band, and line spectra. Having described the production of spectra, and the means by which they are examined, it is necessary to consider the conditions which modify them, and also the laws governing their construction ; much light has been thrown on these subjects by more recent investigations; 1 indeed spectroscopic methods appear emi- nently suited for the elucidation of the molecular structure of matter, since change in a spectrum indicates change in atomic motion. All substances are composed of molecules, consist- ing of similar or dissimilar atoms, the number of which probably varies with the temperature and pressure; the molecules are in a state of active vibration, but, in the case of solids, are maintained in proximity by their mutual attraction, the vibrations being manifested chiefly as heat. The mole- cules of liquids exert sufficient attraction to prevent their complete separation, but those of a gas are independent; their path through space is relatively great and their collisions comparatively few. 1 Kayser, Spectralanalyse (Berlin. 1883). Spectralanalyse in Encyklo- padie der Naturwiss. 32. (Handb. der Physik von Winkelmann. Bres- lau, 1894, p. 419.) W. A. (1891)42. Chem. Ztg. (1892) 16, 593. Kayser and Runge, A. B. A. 1888-1894. Rydberg, Svenska Vetensk. Akad. Handl. (1890) 23. Deslandes, C. r. (1886) 103, 375; (1887) 104, 972. Julius, Ann. de 1'Ecole polyt. de Delft, 1889. Wiillner, W. A. (1874) 8; (1888) 34. E. Wiedermann, W. A. (1878) 5. Schuster, B. A. R. 1880. Lockyer, Studies in Spectrum Analysis (London and New York, 1878). 69 7O SPECTRUM ANALYSIS. In addition to the motion of the molecule as a whole, there is a continuous movement of its parts ; whilst the former r except for the velocity and extent of the free path, is the same for all gases, the motion of the atoms must be different for each kind of molecule, since it will be conditioned by the position, number, and mass of the atoms, by their energy, and by the nature of the collision of different molecules. The vibrations of glowing vapors, which we perceive as light, are conditioned by the vibrations of the atoms, so that change ia these must produce alteration in the spectra. The nature of the relationship between the vibrations of the atoms and the luminiferous ether are unknown, but it may be assumed that the wave-motion of the latter exhibit the vibrations of the former, the number of vibrations of an observed spectral line corresponding with that of the atom itself. Maxwell showed that only the majority of the molecules of a gas are at its mean temperature, the remainder are at all possible intervals above and below this, so that the spectrum produced at a par- ticular temperature is not pure, but a mixture, with those rays predominating which correspond with the mean temperature. Solid bodies have their molecules closely adjacent, the atoms being restrained by external forces from producing their own vibrations, and this proximity causes the production of all possible vibrations if the collisions increase in fre- quency; at low temperatures the vibntions are comparatively slow and produce radiant heat; but as the temperature rises the collisions increase in violence and the vibrations in fre- quency, producing successively the infra-red, the red, the yellow, and so on, until, at the highest temperatures, the ultra-violet is obtained. All solids therefore exhibit contin- uous spectra, containing rays of every possible wave-length ; the same applies to liquids, so far as they can be caused to emit spectra. The circumstances are otherwise with luminous gases and vapors: the intervals between the molecular impacts are relatively long, the characteristic individual vibrations of SPECTRA. 71 the atoms are able to develop, and the corresponding waves appear in the spectrum, which is therefore discontinuous and consists of separate bright regions. Discontinuous spectra are divisible into band and line spectra, formerly termed by Pliicker and Hittorf 1 spectra of the first and second order. The former usually consist of a number of bands, one edge being bright and gradually diminishing almost to darkness in the direction of the other edge; they resemble to some extent illuminated fluted columns, hence the name channelled spectra, which is also applied to them. Observed under high dispersions the channellings are resolved into numerous slender lines, arranged regularly, their proximity being greatest in the brighter regions. Band-spectra are exhibited by compounds, and also by elements at temperatures below that necessary for the production of lines. Line-spectra consist of separate bright lines (slit images) which, if produced by means of a prism, are not perfectly vertical, but are slightly inclined towards the red; they are far less numerous than those in the band-spectra, and appear not to exhibit regularities in position and brightness. The manner in which they change into the very different band- spectra has not been explained ; it is known that the latter are obtained at temperatures intermediate between those required for the production of continuous spectra and line- spectra, and it has been suggested that they are produced by molecular aggregates which would be expected to yield spectra richer in lines than those that could be formed in the presence of fewer atoms. It has long been disputed whether the chief portions of a spectrum are constant when the mole- cules remain the same. Wullner 2 decides in the negative, and holds that, with unchanged molecules, the emission is a function of the temperature, the band-spectrum being pro- 1 P. T. (1865) 155- 2 W. A. (1879) 8; (1888) 34. Ber. Berl. Akad. (1889) 38. Comp. also his " Lehrbuch der Physik." 72 SPECTRUM ANALYSIS. duced at low, and the line-spectrum at high temperatures; these together form the complete spectrum of the particular substance, and the change is continuous. The opposite view is now generally accepted, and has been chiefly developed by Kayser; according to this, so long as the molecules are unchanged their particular vibrations must remain constant, but it does not follow that at any temperature they should exhibit all their possible modes of vibration, and particularly not with equal intensity. It has been repeatedly observed that increase in the violence of impact is correlated with greater intensity in the shorter wave-lengths; the ultra-violet lines become considerably stronger if the arc is used instead of the Bunsen flame, but the longer waves also increase in brightness, lines before too faint to be seen become visible, and there is a general increase throughout the spectrum in the number and brightness of the lines. The spontaneous reversal of the lines is regarded by Kayser as a definite proof of the constancy of the spectrum within each order. Light from a luminous heated gas is absorbed by cooler gas of the same kind, but as the same rays are emitted as are absorbed by the cooler vapor it follows that the wave-length must remain un- changed, although the intensity will be considerably decreased and the original bright lines be replaced by dark ones; since such reversal occurs without alteration between all attainable limits of temperature, about iooo-5OOO, the constancy of the emissions throughout the same range is established. Influence of Temperature and Pressure. It has been stated above that increase in temperature produces greater intensity in the lines within the particular order of spectrum. Increase in the pressure is accompanied by a broadening of the lines; ' this change may be exhibited by all substances in varying degree, and it may occur symmetrically or only towards one side, in the latter case generally towards the least 1 Comp. Schuster, B. A. R. 1880, p. 275. Roscoe and Schuster, Spectrum Analysis (London, 1885), pp. 136, 163. SPECTRA. ^r 73 refrangible end. The hydrogen lines may be extended to such a degree that the spectrum becomes continuous; ' Zollner 2 believed that this was due to the density of the luminous layer; his conclusion was deduced from KirchhorTs law, but it is not in agreement with the observation that a Geissler tube exhibits the same number of sharp lines whether viewed longitudinally or transversely, and that sharp lines are shown by the solar atmosphere and prominences in spite of the enormous thickness of the former. General acceptance is now given to Lockyer and Frankland's 3 view that the increase in breadth is due to greater pressure, although the tempera- ture also exercises some influence; but, in the cases under consideration, a rise in temperature necessarily produces an increase in the pressure. The theoretical explanation of the phenomenon is as follows: so long as the molecules vibrate singly the oscillations occur regularly and at equal intervals, and therefore produce sharp lines, but if other molecules are in close proximity, the vibrations are disturbed by their impact, the frequency of which depends upon the pressure and temperature. 4 Lockyer's Long and Short Lines. Lockyer 5 has devised a method which readily shows the influence of temperature and pressure on a spectrum. The arc or spark is adjusted horizontally to the vertical slit of the spectroscope, and the image thrown on to the slit by means of a lens; a spectrum is thus obtained exhibiting long and short lines of varying breadth: that shown in Fig. 39 is produced by a mixture of calcium and strontium. The image of the slit corresponds with that of a section of the arc, the middle of the image showing the lines in the middle of the arc, those at the sides 1 Frankland, P. R. S. (1868) 16, 416. Wullner, P. A. (1869) 137, 369. 2 P. A. (1871) 142, 88. 1 P. R. S. (1869) 27, 288. 4 Comp. Lippich. P. A. (1870) 139, 465. 5 P. T. (1873) 163, 253, 639. Galitzin, W. A. (1895) 56, 78. 74 SPECTRUM ANALYSIS. SPECTRA. 75 of the latter being shown at the extremities of the image. The luminous vapor is both hotter and denser in the middle than at the sides of the arc; therefore, if the spectrum is influenced by temperature and pressure, the middle of it should differ from its extremities, and this is actually the case. The longer lines are most numerous at the sides, the short ones being confined to the middle; all taper towards their extremities; moreover, the length is not dependent on the brightness of the lines, as the fainter ones may be either short or long. Lockyer considers that the longer lines are produced at lower temperatures, and correspond with the chief lines observed by the ordinary method; the short lines are due to relatively high temperatures, and the expansion in the middle is caused by the greater pressure in the interior of the arc. Influence of Magnetic Current. When the ZMines are produced by means of a Rowland's grating, and a Bunsen burner and sodium chloride, they have been observed by Zeeman ' to widen during the passage of a current, if the burner is placed between the poles of an electromagnet. With an oxy-coal gas-flame they expanded to three or four times the normal width. Similar results were obtained with a lithium line. Interruption of the current produced an immediate reversion to the ordinary state. The widening was also observed in the absorption-lines (reversed lines) produced by sodium vapor, in a porcelain tube placed between the poles and perpendicular to a line joining them. The widening is not due to change in the density of the luminous or absorptive gases, but the observations confirm Lorentz's theory, according to which electrical phenomena are con- ditioned by the position and motion of electrically charged ions, by which also light vibrations are accomplished. Zeeman deduces from this theory the proposition that the broadened 1 Zittungsverl. K. Akad. Wet. Amsterdam (1896-97), pp. 181-242. W. A. Beibl. (1897) 21, 139. Astrophys. J. (1897) 5, 332. P. M. (1897) [5] 43, 226. UNIVERSITY \ 76 SPECTRUM ANALYSTS. spectrum-lines of a light-ray, in the direction of the magnetic current, are subjected to circular polarization, one extremity to the left, the other to the right. If the ray is at right angles to the current, both extremities are linearly polarized, at right angles to its direction. II. ABSORPTION-SPECTRA. Kirchhoff's Law. Fraunhofer, in 1824, observed the coincidence of the yellow sodium lines with the double ZMines of the solar spectrum, and the relationship between the emis- sion and absorption of light had been previously suggested by various workers, 1 but Kirchhoff 2 in 1859 enunciated and estab- lished the law which bears'his name, and which is also known as the " law of exchanges." In order to directly prove the coincidence of the above lines Kirchhoff observed a moderately bright solar spectrum through a sodium flame which was placed before the slit; the dark lines were at once changed to bright ones, but with a very bright solar spectrum the lines were darker than when viewed directly. He then examined the Drummond lime-light through the sodium-flame, and got dark lines in place of the yellow ones, showing that the sodium- flame absorbs the same kind of rays that it emits. The results of these experiments, and certain theoretical consideration^, led him to propo -^d the generalizationfthat the relationship^ between the emissive^and absorptive power of allsubstances Tor light of the same wave-length is identical ~at the same jtemperature. The absorption-speciruTnoTa substance corre- sponds therefore with its emission-spectrum at the same tem- perature and in the same molecular condition. This was proved by Kirchhoff and Bunsen in the case of sodium and 1 Angstrom. P. A. (1853) 94, 141. Foucault, Bull. Soc. philom. de Paris, 1849. A. c. p. (1860) [3] 58, 476. Stokes, P. M. (1860) [4] 20, 20. Balfour Stewart, T. R. S. E. 1858. 2 A. B. A. 1861, p. 64. SPECTRA. 77 other volatile metals, and by Cornu, 1 Liveing and Dewar, 2 and Lockyer 3 for others, including those that are most refractory. Kirchhoff's investigation finally proved the nature and origin of the Fraunhofer lines (comp. Chapter IX). Gases and vapors at low temperatures show absorption-spectra consisting of bands, but at higher temperatures they are composed of lines; as a rule the absorption-spectra of solids and liquids are continuous over a large portion of the field, corresponding with their continuous emission-spectra; the spectroscopic investigation of substances thus becomes possible at a tem- perature below that at which they are luminous. Kirchhoff's law indicates that the luminosity of bodies is due to increase in their temperature. Objection has been made to this by more recent investigators; thus E. Wiede- mann 4 has shown that, apart from the normal evolution of light, causes other than rise in temperature may produce luminosity in a body, and to this luminescence he considers that Kirchhoff's law does not apply. Hittorf 5 and W. v. Siemens 6 have also shown that, up to a temperature of about 2000, gases emit no light, whilst Pringsheim 7 believes that vapors cannot become luminous by increase of temperature alone, but only in consequence of undergoing chemical change. At present it is not possible to say how far these objections are justified; even if correct they do not necessarily invalidate the law of exchanges, which has received support from the theory of resonators. The mechanism of light absorption is as yet far from being completely understood. Influence of Temperature and Physical State. Absorp- tion-spectra are usually observed at low temperatures, thus C. r. (1871) 73, 332. P. R. S. 28. Studies in Spectrum Analysis (London and New York, 1878). W. A. (1888) 34, 446. Ibid. (1879) 7; (1883) 19. Ibid. (1883) 18. Ibid. (1892) 45. 78 SPECTRUM ANALYSIS. readily permitting the determination of the influence of the molecular constitution. The absorption-spectra of iodine in the solid form, in solution, and in the gaseous state are all different; indeed in the latter state it exhibits a line and a band spectrum corresponding with its two emission-spectra. A rise in temperature causes an increase in the absorption within the same order of spectra. H. W. Vogel 1 found that if solutions of organic dyes are volatilized on glass plates, the residue usually exhibits a different spectrum from that of the solution, but the latter remains unchanged if gelatine, glue, starch, or gum arabic is added to the solutions before drying. Stenger 2 states that in the gelatine film the molecular aggregation is the same as in solution, hence the identity of spectrum; in its ordinary solid state the dye is composed of more complex molecules and therefore has a different absorption-spectrum. Influence of the Solvent. Solutions of substances which exhibit absorption-spectra consisting of bands frequently show no regularity in the changes which occur when other solvents are employed. In this connection Kundt 3 has pro- pounded the following rule, which, however, is not of uni- versal application: comparing two colorless solvents which differ considerably in refractive and dispersive power, the one in which these are greater will cause the absorption-bands to approach the red end of the spectrum. Stenger 4 accounts for the exceptions to the above rule by suggesting that the spectrum of a substance is dependent both on its chemical composition and on its molecular state; if the physical mole- cules in the solution are identical with the chemical ones, the body follows Kundt's rule, but solutions frequently contain aggregates composed of a number of chemical molecules. 1 Ber. Berl. Akad. 1878, p. 409. 8 W. A. (1888) 33.583. 3 P. A. Jubelband (1874) p. 615. W. A. (1878) 4, 34. 4 W. A. (1888) 33, 577. SPECTRA. 79 The deviation from, or agreement with the rule may also be due to the varying extent to which the substance in solution undergoes electrolytic dissociation. Influence of Optical Density. 1 The influence of concen- tration and of the thickness of the layer of substance has been already considered in the preceding chapter. The experi- ments of Bunsen and Roscoe 2 show (i) that the quantity of light absorbed by a layer of infinite thickness is proportional to the quantity (intensity) of the incident-rays. (2) The quantity of light absorbed is dependent on the density of the absorbent. The coefficient of absorption, calculated from these data, gives the relationship in intensity between the incident and emergent rays for a layer of unit thickness; in place of this, Bunsen and Roscoe employ the coefficient of extinction, which facilitates the calculation of the concentration from the absorption; the term is applied to the reciprocal of the thick- ness of substance required to reduce the light to one tenth of its original intensity. Fluorescence and Absorption. Absorption of light is connected with phosphorescence and fluorescence. Certain substances become luminous by the action of light; if the luminosity ceases on the withdrawal of the light, they are said to be fluorescent, whilst the term phosphorescent is applied to substances which continue to be luminous after the light is cut off. Hitherto fluorescence has only been observed in the case of liquids, and phosphorescence in that of solids. In accordance with the law of the conservation of energy, the rays causing these phenomena are absorbed ; fluorescent bodies exhibit corresponding absorption-spectra, and, as they absorb the ultra-violet rays more or less completely, they all fluoresce in this region of the spectrum. 1 Comp. O. Knoblauch, W. A. (1891) 43, 738. 8 P. A. (1857) 101, 235. - 80 SPECTfi UM ANALYSIS. III. RELATIONSHIP BETWEEN THE LINES OF AN ELEMENT. Observation of different elements shows that some have lines distributed throughout the whole spectrum field, whilst others exhibit only a few single lines or groups, so regularly arranged as to suggest the idea of a definite relationship. Early investigations led to the conclusion that an acoustical law could be applied to the luminiferous vibrations of the molecules; a string vibrating as a whole gives a fundamental note, but if it vibrates in parts the number of vibrations in the notes produced is 2, 3, 4, etc., times that of the funda- mental note; if this Jaw applies in optics, the wave-length of the different lines of a spectrum must bear a mutual ratio represented by whole numbers. The first attempt to discover such regularities was made by Lecoq de Boisbaudran ' in the case of the nitrogen lines; his conclusions were based on wave-length determinations of insufficient accuracy, aqd were not confirmed by Thalen. Stoney 2 was more successful, and showed that the ratio of the hydrogen lines C : F : h 20 : 27 : 32 ; the subject was further investigated by Stoney and Reynolds, 3 Soret, 4 and others, until the more thorough work of Schuster 6 rendered the theory no longer tenable. He showed that, even when there is absolutely no connection between the lines, the chances are in favor of a harmonic relationship in spectra rich in lines, and, whilst many facts indicate the existence of a mutual relationship between the wave-lengths, the law which it follows is as yet undiscovered. The subject ceased to attract attention for several years until Balmer 6 published a formula which reproduces with wonderful accuracy the posi- C. r. (1869) 69, 694. P. M. (1871) [4] 41, 291. Ibid. (1871) [4] 42, 41. Ibid. [4] 42, 464. B. A. R. 1880. P. R. S. (1881) 31, 337. W. A. (1885) 25, 80. SPECTRA. 8 1 tion of the hydrogen lines. The values are given by the expression ' - 4 n is a whole number between 3 and 15, A a constant, 3645.42 Angstroms according to Cornu's measurements, or 3647.20 taking Ames* more accurate determinations; the possible error in the latter is only 0.1-0.3 Angstroms, and the agree- ment between the observed wave-lengths and those calculated; from the above formula is within these limits. Cornu, 1 simultaneously with Balmer, pointed out that the wave-length of the readily reversible lines of aluminium and thallium bear a definite relation to those of hydrogen, whilst a few years later Deslandres 2 gave a formula for the lines composing the bands of numerous elements. In 1887 Kayser and Runge 3 commenced their investiga- tions, and succeeded in obtaining a formula which reproduces, "series" in the case of a considerable number of elements; Balmer's formula for the hydrogen lines is only a special instance of their more generalized expression. The term "series" is applied to related lines, which are particularly numerous in the spectra of the metals of the alkalis and alkaline earths. Attention had bfeeri^called to these by Live- in% and Dewar * before the publication of Balmer's work. The distance between two consecutive lines decreases with dimin- ishing wave-length, so that the lines asymptotically approach a limit; they applied the term '* harmonic " to such a series of similar groups. 5 Taking th^- refractive index of air as con- 1 C. r. (1885) 100, ii8i. 2 Ibid. (1886) 103, 375; (1887) 104, 972. 3 Ueber die Spectren der Elemente A. B. A. 1888, 1889, 1890, 1891, 1892, 1893. W. A. (1894) 52, 114. Runge, B. A. R. 1888, p. 576. Kayser, Chem. Ztg. (1892) 16, 533. Encyklopadie der Naturw. 32 (Winkelmann's Handb. der Physik. Breslau, 1894) p. 429. 4 P. T. 1883, p. 213, and also previously. 5 Ibid. 1884. 82 SPECTRUM ANALYSIS. stant, a value proportional to the number of waves, i.e., the reciprocal, was used by Kayser and Runge in place of the wave-length; thus modified Balmer's formula becomes - = A + Bn~ 2 , and then T = A + Bn~ 2 + Cn~*. A A This expression gives only an approximation; probably the number of waves is only a function of n which, in the nega- tive power, admits of the development of a rapidly conver- gent series; of these the first three terms are sufficient to give their values with remarkable accuracy. Kayser and Runge then extended their investigations so as to elucidate the fol- lowing questions: the applicability of the formulae in the case of measurements of the highest possible degree of accuracy; whether lines of wave-lengths indicated by the formulae really exist; can all the lines of every element be reduced to series? can a relationship be shown between the constants of the formulae of different elements ? The investigators' objects could not be attained by the tise of the older wave-length measurements, partly on account of their inaccuracy, partly because they did not include the ultra-violet, so that it became necessary to re-examine the spectra of the elements with the highest possible degree of accuracy. The largest number of series represented by the above formulae are exhibited by the spectra of the members of Mendeleeffs first three groups. The metals sodium, potassium, rubidium, caesium, copper, silver, aluminium, indium, and thallium each have two series in which B and C are identical and A differs; two such series may therefore be regarded as a series of pairs of lines, each pair having the same difference in vibration. Probably all elements have two such series of pairs. The first series contains strong, ill- defined lines, and is termed the first "subseries " ; the second series contains well-defined fainter lines, sometimes broaden- SPECTRA. 83 ing out towards the red : they form the second subseries. The two subseries were not observed in the spectra of rubidium and caesium; lithium exhibits both series, but they consist of single lines instead of pairs. The difference in wave in both series is practically identical in the case of each element, and bears a relationship to its atomic weight. The alkali metals have a third series which includes the strongest and most readily reversible lines of the whole spectrum, and is called the " principal series " ; in the spectrum of lithium it consists of single lines, in the other metals of pairs; these are closely adjacent in the case of sodium, but with increasing atomic weight the separation becomes greater, whilst the entire series gradually approaches the least refrangible portion of the spectrum. In each pair the stronger line has the smaller wave-length; this was already known to be true of the sodium jtMines. Within each principal series the difference in the number of waves between the pairs is approximately inversely proportional to the fourth power of the ordinal number. The largest positive value given by the formulae for all series hitherto observed corresponds with the ordinal number 3; the lines where n =. 3 are comparable with fundamental notes, since they represent the longest possible waves, exactly as exhibited in Balmer's formula for the hydrogen lines. The spectra of copper, silver, and gold do not show such striking regularities as those of the alkali metals, which appear all to be arranged on one plan. By analogy with the order observed in other spectra the existence in the spectra of copper and silver of both subseries of pairs can be demon- strated ; but this is not so with gold, possibly because the series become fainter as the atomic weight increases. Magnesium, calcium, and strontium, amongst the alkaline earths, have spectra with two subseries consisting, not of pairs, but of triplets with a constant difference of wave: as the atomic weight increases the series diminish in intensity and approach the red end of the spectrum. This probably ex- 84 SPECTRUM ANALYSIS. plains why no series could be found in the case of barium r the last element in this group. The spectra of zinc, cadmium, and mercury also exhibit two subseries of triplets, but scarcely half the total number of lines is included in the series. Only a few of the elements in Mendeleeff's fourth and fifth groups have been available for Kayser and Runge's in- vestigations, which have been confined to tin and lead in the former, and to arsenic, antimony, and bismuth in the latter; the regularities found in the members of the first three groups could not be detected in these. Each spectrum is character- ized by a large group of lines which are repeated in such a way that, by the introduction of a constant, the number of waves of one group may be deduced from that of another, but the lines do not permit of arrangement into series, and their appearance gives no clue as to their possible interrela- tionship. It is not surprising that all the lines of a spectrum do not fall into series, for in order to compare different elements they should be investigated under the same relative conditions, and not at the same temperature. Failing accu- rate knowledge of the temperatures at which the elements would be in a uniform molecular condition, it may be assumed that those of high, melting and boiling point would require a much hotter flame than the more volatile ones; consequently if an arc lamp, giving a temperature of 3OOO-35OO, is required in order to produce the complete series in the case* of the readily fusible alkali metals, it follows that, with the other elements, the higher the melting-point the less characteristic will the series be. Working independently of Kayser and Runge, Rydberg 1 simultaneously adopted the same view of the structure of 1 C. r. (1890) 11O, 394. Zeitschr. physikal. Chern. (1890) 5, 227. Svenska Vetenskap. Akad. Handlingar Stockholm (1890), 23, No. n. W. A. (1893) 50, 629; (1894) 52, 119. \ SPECTRA. 85 line-spectra; he employed the number of waves instead of the wave-length, and his investigations of the members of the first three groups of the periodic system led him to conclude that the "long" lines form pairs or triplets which, in the case of each element, are characterized by a constant difference (v) in the number of waves of the components. In each group of elements this value increases in a ratio somewhat exceed- ing the square of the atomic weight. The triplets occur in the first and third groups, the valency of which is odd; the components of the double lines form series, the members being functions of consecutive whole numbers; each series can be approximately reproduced by the expression where n is the number of waves, m any whole number, the ordinal number of the member, and N Q = 109721.6, a con- stant applicable to all the series of every element, and which is obtained from Balmer's formula; n and /* are constants peculiar to each series, n being the limit which the number of waves n approaches if m is infinite. Like Kayser and Runge, Rydberg distinguishes three kinds of series, "nebu- lous," "sharp," and ''principal"; the first two are composed of pairs or triplets, so that the elements of the first and third groups have four different series of these two kinds, and the elements of the second group have six; they are termed the first, second, and third nebulous or sharp series; the lines of the first series of either kind are the strongest and least refractive. In the case of the elements in group one, the principal series contains the strongest lines of the spectrum, the nebulous series are next in order, and the sharp series the faintest; in both the separate groups and series the intensity of the light decreases as the ordinal number rises. The different series of an element are sufficiently related 86 SPECTRUM ANALYSIS. to show that they all belong to one system of waves; the series of the same group, nebulous or sharp, have the same value for //; the difference of the n value is equal to v or v l and ^ 2 ; the series of the same order, first, second, or third, have the same value for n in the different groups, but differ in that for yw. The wave-length and the corresponding number of waves, the values of the constants v, n , and p of the corre- sponding series in the various elements, are periodic functions of the atomic weight; the periodic difference in the constants permits of the calculation of the spectrum of an element if the spectra of the elements adjacent to it in the periodic system are known. Rydberg's investigations have strengthened the arguments in favor of a single system of waves, and indicate the possi- bility of representing all the lines of a spectrum by a single formula, but they are opposed to the idea of a mixed spec- trum such as would be produced by molecules at varying temperatures. He considers it probable that each element possesses only a single spectrum, and that the intensity of the series and of the special lines varies with the temperature and density of the luminous gas, in a manner similar to the changes in the overtones of a bell. The arrangement of band-spectra suggests, even more strongly than line-spectra, the possibility of their structure conforming to definite rules; Lecoq de Boisbaudran ' and Thalen 2 pointed out certain regularities, but these did not permit of the deduction of a law which was first formulated by Deslandres. 3 The lines of a band form series of similar lines, the series being connected in such a manner that, in each one, the distances between two consecutive lines are approximately in arithmetical progres- sion. If the edge of a band is designated by o and the follow- 1 C. r. (1869) 69. 2 Svenska. Vetensk. Akad. Handl. (1869) 8. 3 C. r. (1886) 103, 375; (1887) 104, 972. A. c. p. (1888) [6] 15. J. de Phys. (1890) [2] 10, 276. SPECTRA. 87 ing lines by the succeeding numbers I, 2, 3, . . . n, then the number of waves of the nth line is given by the formula - n "n where a is the number of waves of the edge, and b the differ- ence between this and the number of waves of the first line. The different bands of a spectrum are related in such a manner that the first, second, etc., edges of all are represented by the expression I = A + B + C n \ corresponding with that representing the lines of a series; A, By and C are constants, and n progressive numbers. The absolute validity of these laws is questioned by Kayser and: Runge, 1 but maintained by Deslandres. Theoretical articles on the origin of lines, pairs, etc., have been published by Lecoq de Boisbaudran, 2 Stoney, 3 Julius, 4 and v. Kovesligethy.* IV. RELATIONSHIP BETWEEN THE SPECTRA OF DIFFERENT ELEMENTS. Attention was first directed to this subject by Lecoq de Boisbaudran 6 in 1869; he pointed out the similarities in the structure of the spectra of potassium, rubidium, and caesium, and, applying the term *' homologous " to certain analogous lines in each, he concluded that in the case of the metals of the alkalies and alkaline earths the spectra -approximate the red as the atomic weight increases. This has been confirmed A. B. A. 1889. C. r. (1869) 69, 445, 606, 657. P. M. (1871) [4] 41. Trans. Dubl. Soc. (1891) [2] 4. P. M. (1892) [5] 33. Ann. de 1'ecole polyt. de Delft (1889), 5. Theor. Spectralanalyse (Halle, 1890). C. r. (1889) 69, 610. SPECTRUM ANALYSIS. by later and more exact measurements. He ' subsequently employed these homologous lines for the calculation of the atomic weights of gallium and germanium, which had not then been determined; his method was based upon the rule which he had enunciated, that within the groups of the periodic system the variation in the increase of the atomic weight is proportional to that of the increase in the wave-length of the homologous lines. The following is an example of the method of calculation: Atomic Weight. Difference. Difference. Variation. Mean Wave- length of Two Lines. Diff. Variation. Si.. Ge.. Sn.. Al.. Ga.. In.. 28.0 n 118.0 27-5 69.9 II3-5 90.0 Between Si and Sn 42.4 43-6 1.2 1.2 _ 2.8302 4010 4453 5077 3952 4101 4306 443 624 149 205 40.51 100 IOO 42.4 100 The fraction - : means that in order to obtain 43.6, i oo 2.8302 per cent of the difference 42.4 must be added; the variation (x) for the group Si, Ge, Sn is obtained from the ratio 37.584 : 2.8302 :: 40.51 : x 3.051 per cent; the in- oo crease (y) in atomic weight from Si to Ge = = 44.32, 2.03051 and therefore the atomic weight (n) of germanium = 72.32. Kayser 3 considers that, whilst the above rule perhaps contains a nucleus of truth, it is at present not applicable, and requires a knowledge of atomic weights considerably exceeding in accuracy almost all the current values. In consequence of this Ames 3 was unable to apply the rule to magnesium, zinc, 1 C. r. 86, 943; (1886) 102, 1291. Ber. 19, 4790. 9 Spectralanalyse, in Encyklopadie der Naturw. 32 (Winkelmann, Physik. Breslau, 1894) p. 440. 3 P. M. (1890) [5] 30. SPECTRA. 89 and cadmium, although his fundamental homologous lines were correctly selected; since the selection of homologous lines is to some extent arbitrary, the close agreement in the calcu- lated values for gallium and germanium must have been partly due to chance. Ditte, 1 Troost and Hautefeuille, 2 Ciamician, 3 Hartley, 4 Ames, 6 and Griinwald 8 have also investigated the interrelationship of the spectra of various elements, but Ames' work alone has proved to be of permanent value. He measured the wave-length of the triplets in the spectra of zinc and cadmium, and calculated the differences in the number of waves between the third lines of the triplets; they decrease from triplet to triplet, are nearly identical for each element, and prove the lines to be true homologues. In the more strongly nebulous series the values for zinc are 581, 263, 141, 84, and for cadmium 587, 264, and 84. The investigations of Kayser and Runge, and of Rydberg have thrown most light on the relation between the spectra of different elements. Their method consists in a combina- tion of calculation and observation, their own exact remeas- urements of spectrum-lines being utilized by the first two observers. The relationship of the spectra of different ele- ments follows from the law already stated which expresses the connection between the lines of a single element. In spectra of similar structure the homologous lines are those with identical ordinal numbers. The work hitherto completed shows that the spectra fall into the same groups as the ele- ments. In the case of the first three groups of the periodic C. r. (1871) 72, 620. . Ibid. Wien. Ber. (1878) 78, 767; (1880) 82 [2]. Jour. Chem. Soc. London (1882) 84; 1883, 390. P. M. (1890) [5] 30, 33. Astr. Nachr. (1887) 117. Wien. Ber. (1887) 96 [2]; (1889) 97 [2]; (1890) 98 [2]. Wien. Anz. 1890. Comp. also Ames, N. (1888) 38. Kayser, Chem. Ztg. (1889) 13; (1890) 14. Runge, P. M. (1890) [5] 30. Grim- wald, Chem. Ztg. (1890) 14. SPECTRUM ANALYSIS. s->m 's- /a "i s- K n I - S-'N : SPECTRA. 91 system the subdivisions are also well marked, so that the following classification may be made: 1. Lithium, sodium, potassium, rubidium, caesium. 2. Copper, silver. 3. Magnesium, calcium, strontium. 4. Zinc, cadmium, mercury. 5. Aluminium, indium, thallium. In each of the above groups the spectrum approaches the red as the atomic weight increases, but in passing from group to group it approximates towards the violet. The systematic representation of these spectra as given by Kayser and Runge is shown in Fig. 40; the values given are the number of vibrations, and, for the sake of clearness, only the first lines of pairs and triplets which have been actually observed are shown ; the figures opposite to the lines are their ordinal numbers. The interrelationship of spectra and atomic weights has been already referred to: it may be briefly expressed by saying that the breadth of pairs and triplets, measured by the difference in the number of their waves, is approximately proportional to the square of the atomic weight. CHAPTER VII. SPECTRA OF THE ELEMENTS. AN accurate knowledge of spectra is of the greatest im- portance for any application of spectrum analysis ; the standard of measurement is the wave-length in air, at medium tern- o peratures, under a pressure of 760 mm. expressed in Ang- strom's units (A) or tenths (/*/*). Until recently all observa- tions were based on Angstrom's wave-length determinations, and on his drawings of the solar spectrum (Spectra normal du Soleil 1 ); this scale was universally employed during twenty years, but after Angstrom's death it was shown by Thalen 2 to be inaccurate in consequence of Angstrom having used a reputed metre measuring-rod less than one metre in length. New determinations of absolute wave-lengths have been subsequently made by Miiller and Kempf, 3 Kurlbaum, 4 Peirce, 5 and Bell; 6 of these the values for the Z^-line of Peirce and of Bell agree exactly, and that of Muller and Kempf very closely; the latter is used as the basis of the Potsdam system. Since the relative values are often of greater importance for spectroscopic purposes than the abso- lute ones, Rowland 7 combined the various numbers as shown below, and employed the mean value as the foundation of his 1 Recherches sur le spectre du soleil (Upsala, 1868). 2 Sur le spectre de fer. N. A. S. U. 1884 [3]. 3 Publicat. des Astrophysikal. Obs. zu Potsdam (1886), 5. 4 W. A. (1888) 33, 159, 381. 5 Sillim. Journ. [3] 18, 51. 6 P. M. (1888) [5] 25, 245, 350. 7 Astronomy and Astrophysics (1893), 12, 321. P. M. (1894) [5] 36, 49. A list of the standard wave-lengths is given in the chapter on the solar spectrum together with references to Rowland's latest publications on the subject. 92 SPECTRA OF THE ELEMENTS. 93 solar atlas, and standard wave-lengths obtained by the coin- cidence method; as this admits of a degree of exactitude (o.oiA) otherwise unattainable, all recent measurements have been based on his scale. Relative Weights. Observer. A- I Angstrom corrected by Thalen 5805.81 2 Miiller and Kempf . . 5806.25 2 58o5 . QO 5806 2O IO Bell 5806. 2O 5806 156 The wave-lengths of the spectrum lines given in the suc- ceeding pages are all based on Rowland's system ; this has entailed a recalculation, by the use of Watts' tables, 1 of all measurements published before 1889, and also of certain others. The object of the tables was to exhibit the relation- ship between Angstrom's and Cornu's solar atlases and Rowland's. To reduce Angstrom's scale to his own system of wave-lengths Rowland multiplies the values by the factor 1. 00016. Such recalculations are open to objections, but these are overruled by the great inconvenience of wave-lengths determined by two different scales, particularly when they refer to the same element; moreover the table permits of the revised values being reconverted into the original ones. The accuracy of the older measurements should not be overrated, it falls far short of that attainable by the use of the grating and photographic appliances, such as have been used by Kayser and Runge, Liveing and Dewar, Hartley and Adeney, Hasselberg, Ames, Trowbridge, and others. Most of the older measurements, with the refer- ences, have been taken from Watts' '* Index of Spectra,"* 1 B. A. R. (1890). (London, 1891), p. 224. * Manchester, 1889. Continuations are given in the B. A. R. 94 SPE CTK UM A NA L YS1S. TABLE FOR THE REDUCTION OF ANGSTROM'S AND CORNU'S WAVE-LENGTHS TO ROWLAND'S VALUES, DERIVED FROM THE UNIT D\ 5896.156. Wave-length. Corr. Wave-length. Corr. Above 6930 + 1-7 From 4970 to 4935 4-1.0 From 6930 to 6880 1.6 " 4935 " 4865 0.9 " 6880 " 6820 1.5 " 4865 " 4740 I.O 6820 " 6800 1.4 " 4740 44 4650 0.9 " 6800 " 6765 1-3 " 4650 44 4470 0.8 4 6765 " 6720 1.2 " 4470 44 438o 0.7 " 6720 " 6660 I.I " 438o " 4170 0.6 " 6660 " 6230 I.O " 4170 " 4130 0.7 44 6230 " 6180 0.9 " 4130 " 4100 0.8 44 6180 " 6i55 I.O 44 4100 ' ' 4060 0.7 n 6i55 " 6135 I.I " 4060 4 ' 4040 0.6 6i35 " 6130 I.O 44 4040 " 3850 0.7 44 6130 " 6110 0.9 " 3850 ' 3730 0.6 " 6110 " 6080 1.0 44 3730 " 3720 0-5 4 6080 " 6060 1. 1 " 3720 ' ' 3660 0.4 44 6060 " 6000 I.O " 3660 14 3640 0.8 44 6000 ' ' 5970 0.9 " 3640 44 3620 0.6 1C 5970 " 5810 I.O 3620 " 3530 0.8 " 5810 578o 0.9 " 3530 4 348o 0.6 " 578o " 5610 I.O 44 348o ' 3470 0.8 44 5610 " 5540 I.I 44 3470 " 3440 0.7 II 5540 " 5485 I.O 41 3440 " 3420 i.i 14 5485 " 5435 0.9 3420 " 336o i-7 II 5435 " 5350 I.O 44 3360 44 3330 2.5 44 5350 " 5335 0.9 44 3330 44 3290 2.2 14 5335 " 5325 I.O 41 3290 44 3280 2.O " 5325 " 5300 0.9 44 3280 44 3240 1.9 4< 5300 " 5175 I.O 44 3240 " 3220 1.8 I 5175 " 5150 0.9 44 3220 44 3190 0.8 11 5150 " 4990 0.8 44 3190 44 3160 0.4 4990 " 4970 0.9 which gives a fairly complete list of the measurements of line- spectra made before its appearance. In the following pages the lines of each spectrum or portion of a spectrum are all from one series of measurements, and that the newest or most trustwothy; only the brighter lines have been included. SPECTRA OF THE ELEMENTS. 95 in general those between I and 3 on the German scale of brightness, in which the brightest is I and the faintest 6. In the English scale the brightness increases from I to 10, so that the lines included are those between 6 and 10, but where the fainter lines are characteristic they have also been given. The lines are not provided with intensity-scale numbers which are only of value in the case of closely adjacent lines, whilst their assignment is always somewhat arbitrary, but in order to. facilitate orientation the specially bright lines have been printed in bolder type. The arrangement of the tables follows the ordinary plan: double or triple lines are enclosed in parentheses; bands are indicated by b, those sharply bounded at the red end and shading gradually towards the violet are distinguished by b r , those showing the opposite behavior by b v . Lines which are usually designated by a number or letter, such as the Z^-line, have these enclosed in brackets, and prefixed to the wave-length; measurements are given to as many places of decimals as are required by the accuracy of the observation. Listing's 1 scale is used for the classification of the lines according to color; it runs as follows: .... to 7230 infra-red. 5850 to 5750 yellow. 454 to 4240 indigo. 7230 " 6470 red. 5750 " 4920 green. 4240 " 3970 violet. 6470 " 5850 orange. 4920 " 4550 blue. 3970 " ultra-violet. The delicacy of spectrum reactions has been determined by Kirchhoff and Bunsen for certain flame-spectra, and by Cappel for a series of spark-spectra; the number in the table below gives that fraction of a milligram of pure substance that could be detected. KIRCHHOFF AND BUNSEN. S Barium chlorate 1000 Potassium chlorate 1000 Caesium chloride 20000 Rubidium chloride 5000 Calcium chloride 16666 Sodium chlorate 3009000 Lithium chlorate nun Strontium chloride 16666 1 P. A. (1868) 131, 564. 2 Ibid. (1860) 110, 161. 9 6 SPECTRUM ANALYSIS. Cappel's 1 results are as follows: Bismuth 7OOOO Indium . . . 9OOOO Potassium iSoOO Iron . . 26OOO 4OOO Lead 2OOOO Calcium. . . . Chromium. . Cobalt . I 0000000 4000000 ISOOO Lithium. . . . Magnesium. Manganese. 4OOOOOOO 5OOOOO 200000 Strontium .. Thallium. . . Tin .. I 00000000 80000000 i7orx> Copper 20000 Mercury... 10000 Zinc, 600000- ALUMINIUM. The visible portion of the spark-spectrum of aluminium has been investigated by Kirchhoff, 2 Thalen, 3 and Lecoq de Boisbaudran, 4 and the ultra-violet region by Hartley and Adeney, and Cornu ; 5 the latter gave a graphic representation of the lines of shortest wave-length, together with a formula from which Julius 'was enabled to calculate the wave-length. The arc-spectrum has been measured by Liveing and Dewar, 7 and more recently by Kayser and Runge, 8 who were unable to detect a single line in the visible spectrum, although the bands of alumina were always visible. These have been investigated by Hasselberg, and are given in a separate table below. Arc and spark spectra: 5723.5* 5696.5* 5057.4* 4662.9* 3961.68 3944 26 3092.95 3092.84 3082.27 3066.28 3064.42 3060.04 3057.26 3054-81 3050.19 2660.49 2652.56 2575.20 2568.08 23/8.52 2373.23 2367.16 2269.20 2263.52 2210.15 2204.73 2174.13 2168.87 2150.69 2145-48 1989.90 1935.25 1862.20 1854.09 1 P. A. (1870) 139, 628. 2 A. B. A. 1861. 3 N. A. S. U. (1868) [3] 6. 4 Spectres lumineux (Paris, 1874). 6 Spectre normal du Soleil (Paris, 1881). C. r. (1885) 100, 1181. 6 Naturk. Verh. d. Akad. v. Wetensch. Amsterdam (1888), 26. 7 P. R. S. 28, 367. P. T. (1883) 174, 220. 8 A. B. A. 1892. Runge, W. A. (1895) 55, 44. See also E. Becquerel, C. r. 96, 1218; 97, 72. * Visible only in the spark spectrum. (Thalen.) SPECTRA OF THE ELEMENTS. 97 ALUMINIUM OXIDE. Arc-spectrum : Group 5210 5079: 5162.05 5156. 45 5155.42 5147' 93 5143.27 5I43-08 5123. 79 (5123.57 5123.47) 5102.84 5102 .32 5079.52 Group 5041 4842: 49I4-35 4909. 55 4908.21 4906.71 4906.52 4906.07 495- 22 4905.04 4904. 84 4903.72 4903. 54 4899.16 4895.20 4895. 00 4892.32 4890. 44 4888.57 4888.41 4887.79 4886.08 4885. 87 4883.45 4882. 43 4882.24 4881. 25 4880.07 4879- 9i 4878. 90 4878.79 4877. 75 4876.64 4876. 56 4875.46 4873. 50 4873. 35 4872.46 4872. 29 4871.48 4870.46 4869.45 4868. 42 4867. 48 4866.54 4863. 09 4862.77 4842.44 Group 4842 4648: 4810.16 4809. 80 4766.75 4766. 53 4760.32 4752. 53 4752. 27 4749-19 (4745- 17 4744.95) 4742. 56 (4736.08 4735- 94) 4727. 40 (4719.41 4719. 29) 4715.45 4711. 98 4711.81 (4707. 53 4707. 26) (4706.26 4706. 17) (4706.01 4705. 89) 4699.00 4697. 90 (4695. 30 4694.78) 4689. 77 4672.15 4658. 68 4655.34 4648. 14 Group 4648 4471. 4593-97 4570.44 4557.84 4547- 33 4543-23 4537. 69 4534- 24 (4523.45 4522. 86) 4516.54 45II- 38 4494.22 4478. 64 4470. 63 ANTIMONY. The spark-spectrum is obtained either by the use of the metal or of a concentrated solution of the chloride; it has been measured by Kirchhoff, 8 Huggins, 3 Thalen, 4 and Hartley and Adeney. 6 The arc-spectrum, which differs from that of the spark, has been investigated by Liveing and Dewar, 8 and the portion commencing at A = 643;^ by Kayser and Runge. 7 1 Hasselberg, K. Svenska Vetensk. Akad. Handl. (1892) 24, No. 15. Lecoq de Boisbaudran. Spectres lumineux (Paris, 1874). Thalen, Upsal. Universit. Arsskrift. 1866. Lockyer, P. T. 163, 658. A. B. A. 1861. P. T. (1864) 154, 139. N. A. S. U. (1868) [3] 6. P. T. (1884) 175, 126. Ibid. (1883) 174, 221. A. B. A. 1893. Also Lockyer, P. T. (1873) 163, 369. Lecoq de Bois- baudran, Spectres lumineux (Paris, 1874). 98 SPECTR UM ANA L YSIS. In the visible portion of the field the lines are feeble and ill- defined, and are not present in the spark-spectrum, the lines of which are absent from the arc-spectrum. According to Lockyer and Roberts ' antimony vapor produces a continuous absorption-spectrum in the blue. Arc and spark spectra: 6302.8* 5568.25 3598.6! 33389t 2QI3.lt 2670.73 2506.9! 2262.55 6129.7* 4949-7* 3566.8! 3305-4t 2890.7! 2652.70 2445-59 2179-33 6079.2* 4878.6* 3559-9t 3267.60 2878.01 2631.6! 2383.7^ 2175.99 6004.7* 4592.4* 3505.2f 3232.61 2790.0! 2616.7! 2373-78 2098.47 5910.1* 4352.6* 3499- if 3029.91 2770.04 2612.40 2360.60 2068.54 5894.6* 4265.6* 3474-7! 2980.2! 2719.00 2598.16 2311.60 5639-1* 3739-6! 3427.0! 2965.6! 2682.86 2528.60 2306.56 ARGON. Argon was isolated from the atmosphere in 1894 by Rayleigh and Ramsay. 2 The pure gas, in a Geissler tube, exhibits several lineal spectra depending on the pressure in the tube and on the nature of the electric discharge. Crookes 3 discovered two of these, and from the predominant color of their light termed them the red and the blue spectrum respectively. Eder and Valenta 4 have observed a third spectrum, which they term the white. It is produced by the use of very large condensers in conjunction with a powerful induction-coil and a strong current. In these circumstances, under a pressure of 15-20 mm., white light is emitted from the capillary. Spectroscopically the light is peculiar; the majority of the lines become widened, and few remain sharp, many coincide with lines in the blue and red spectra, but 1 P. R. S. (1875) 23, 344. 2 P. T. (1895) 186, 221. 3 C. N. (1895) 72, 66, 99. 4 Sitzungsbericht d. Wiener Akad. Mathem.-Naturw. (1895) 404. Denk- schr. d. Wiener Akad. (1896) 64. * Visible only in the spark-spectrum. (Thalen.) ! Visible only in the spark-spectrum. (Hartley and Adeney.) SPECTRA OF THE ELEMENTS. 99 o certain groups are displaced from 0.5-1 Angstrom towards the red. At present Eder and Valenta are unable to suggest the cause of this partial displacement, but it appears to be connected with the pressure and temperature of the gas, and with the nature of the electric discharge. There is still some doubt as to whether argon is an element or a mixture of two. Dewar, and subsequently Berthelot have suggested that it is an allotropic modification of nitrogen, but later work does not lend confirmation to this view. The gas obtained from cleveite, which was formerly supposed to be nitrogen, has been shown by Ramsay to exhibit all the lines of atmospheric argon together with several others including the Z> 3 -line f helium ; but atmospheric argon contains at least three bright lines in the violet which are not shown by the gas from cleveite; hence Ramsay concludes that atmospheric argon is probably a mixture. Berthelot l obtained a fluorescent spec- trum by the action of a moderately strong induction-current on a mixture of argon, benzene vapor, and mercury in a Geissler tube; the spectrum differs from that given by any other gas, and the yellow and green rays were perfectly visible in the spectroscope in full daylight. He considers that the spectrum is that of a compound of argon and mercury with the constituents of benzene, but Dorn and Erdmann 2 found that some of the lines were those of mercury and nitrogen. Eder and Valenta 3 have photographed the argon spectrum between A. = 5060 and 332O/*/-*, using a powerful concave grating, and Kayser 4 has published a preliminary list of the lines in the blue spectrum, the gas being obtained from the atmosphere; the lines observed are not given in Rowland's Atlas and reproductions of the Fraunhofer lines. 1 C. r. (1895) 120, 662, 797, 1049, 1386; (1897) 124,~ 113. 2 Lieb. Ann. (1895) 287, 230. 3 Wiener Akadem. Anzeiger (1895), No. 21. 4 C. N. (1895) 72, 99. Sitzungsber. d. ,Ber,k Akad..(i896) 24. See also Newall, C. N. 71, 115. 100 SPECTR UM ANAL YSIS. Trowbridge and Richards ' find that the oscillatory dis- charge of the condenser is an important factor in producing the blue spectrum of argon. The pure red spectrum is obtained if the tube is connected with the terminals of an electric machine; but if the spark-gap is interposed, the spec- trum changes at once to blue. Red spectrum: 7723.4 7635.6 7515-4 7383-9 7066.6 6964.8 6415.2 6031.5 5739.87 5651.03 5607.44 5597.89 5572.87 5559-02 5506.42 5496.16 5451.95 5421.68 5221.65 5187.47 5162.59 4888.21 4702.40 4628.60 4596.30 4522.49 4510.90 4345.27 4335.42 4333-65 4300.18 4272.29 4266.44 4259.50 4251.27 4200.75 4198.40 (4191.02 4190.85) 4182.03 4164.36 4158.65 4152.97 4054.65 4046.04 4044.52 3949.08 3947-75 3894.78 3834-83 3781.07 3680.30 3678.43 3649.99 3634.64 3632.82 3606.69 3588.64 3567.88 3564.54 3563.50 3554.48 3461.23 3394.03 3034.7 3021.9 2967.3 2614.6 2516.3 2478.65 Blue spectrum: 6644.2 6059.5 6043.0 6031.5 5651.03 5607.44 5559 02 5496.16 5287.24 5166.03 5145.57 5142.20 5062.35 5017.46 5009.63 4965.38 4933-49 4880.14 4866.14 4847.94 4806.17 4765.04 4736.03 4658.04 4637.35 4609.73 4590.05 4579-53 4545.26 4503.15 4481.99 4426.16 4401.19 4400.25 4379-79 4371.46 4370.92 4352.40 4348.11 433L3I 4283.03 4277.65 4266.44 4237.34 4228.27 4222.76 4182.97* 4179.45 417^.58 4175-25* 4174.20* 4172.95* 4172.05* 4156.30 4I3L95 4113.04 4104.10 4082.59 4079.80 4076.85 4072.58 4072. 18 4053-12 4043.04 4038.99 4035.58 4033.99 4013.97 3992.17 3979-57 3974.70 3946.20 3944.50 3932.71 393L32 3928.78 3925.93 3914.93 3911.69 3907.80 3892.15 3891.53 3880.46 3875.40 3872.26 3868.68 3850.70 3845.51 3841-63 3830.58 3826.92 3809.58 3808.72 3803.38 3799-65 3795.56* 3786.60 3781.07 3766.30 3765.48 3763-76 3753.6o 3738.04 3734-70 3729.52 3720.61 3718.39 3717.36 3660 70 3656.26 3655.52 3651.04 3640.00 3637.25 3622.31 358864 3582.54 3581.82 3576.80 3565-20 3561.20 3559.69 3548.69 3546.03 3545.78 3535-53 3522.14 3521.46 3520.15 35I4.53 3509.93 3491.71 3480.69 3478.42 3476.96 3464.33 3454-30 3421.80 3391-86* 3388.65 1 Amer. Jour. Sci. 1897 [4] 3, 15. P. M. 43, 77- See also Friedlander, Zeit. f. phys. Chem. (1896) 19, 662. * Visible only by the use of powerful condensers, otherwise absent from the normal blue spectrum. SPECTRA OF THE ELEMENTS. 1OI 3376.61 3351-10 3307-37 3301.97 3293-82 3285.91 3281.83 3263.71 3204.49 3181.26 3169.88 3161.64 3139.26 3093.57 3029.10 2979-35 2955.67 2943.17 2924.92 2896.97 2891.87 2866.0 2806.3 2769.7 2753-9 2744-88 2732.67 2708.40 2647.6 2562.3 2544.8 2534.8 2516.8 2515-6 2500.4 2491.0 2480.9 2479.2 2454-5 2438.8 2415.7 2395-7 2364.2 2350.6 2344-4 2337.8 233L7 2316.5 2314.0 2309.4 2282.6 2252.4 2243.7 2234-7 2219.9 2175.6 2171.5 2165.8 2130.6 2050.5 White spectrum : 5306.04 5287.24 5166.03 5145-57 5142.20 5062.35 5017.46 5009.63 4972.40 4965.38 4933-49 4888.88 4880.14 4867.72 4847.94 4806.17 4765.04 4736.03 4727-00 4658.04 4609.75 4590.04 4579.53 4545.26 4481.99 4430.35 4426.16 4401.19 4400.25 4379-79 4371.46 4370.92 4352.40 4348.11 4332.20 4331-31 4278.02 4266.44 4228.27 4104.93 4072.3 4013.97 3933-40 3928.78 3892 15 3869.50 3850.70 3827-67 3781.58 3766.21 3729.52 3589.11 3582.79 3577-27 3561.50 3560.15 3546.58 3514.98 3510.26 349I-7I 3477.38 3388.94 3377.J8 335i.8o 3294.58 ARSENIC. The spark-spectrum of arsenic is obtained by the use of the vapor of the element, or of the chloride contained in a Geissler tube. The arc-spectrum differs from that of the spark, and exhibits no lines in the visible field; the portion from 6ooyw// onwards has been photographed by Kayser and Runge;' between 3OO/f/u and 2OO/^u the lines are numerous though not very strong, but they are often observed, showing the wide distribution of arsenic, and its frequent occurrence as an impurity ; indeed the lines \ = 2349 and 2288, which are the strongest, are rarely absent from any spectrum of a carbon arc. Lockyer 2 and Ciamician 3 have described a channelled absorption-spectrum. 1 A. B. A. 1893. 2 C. r. (1874) 78, 1790. a Wien. Ber. 76, 499; 78, 867; 82,425. See also Thalen, N. A. S. U. (1868) [3] 6. Hartley and Adeney, P. T. (1884) 175, 124. Kirchhoff, A. B. A. 1861. Huggins, P. T. (1864) 154, 139. Plucker and Hittorf, P. T. 155, i. Ditte, C. r. (1871) 73, 738. Huntingdon, Sillim. Jour. (1881) 22, 214. 102 SPECTRUM ANALYSIS. Arc and spark spectra: 6170.7* 6111.2* 5652.1* 5559-2* 5499-1* 5332-1* 4494-7! 4467.0! 4459-4t 4431- 7t 4036.7t 3949.2! 393i-4t 3922.3! 3825.1! 3785-0! 3119.69 3075.44 3057.7! 3053-0! 3032.96 2991.11 2898.83 2860.54 2830.2! 2780.30 2745.09 2601.2! 2528.3! 2526.4! 2492.98 2456.61 2437.30 2381.28 2370.85 2369.75 234992 2288.19 2271.46 2228.77 2165.64 2157.1! 2148.2! 21.14.21 2133.92 2113.14 2067.26 2009.31 BARIUM. The spark-spectrum of barium has been investigated by Kirchhoff, 1 Huggins, 2 Thalen, 3 and Lecoq de Boisbaudran; 4 the arc-spectrum by Lockyer, 5 Liveing and Dewar/ and, most accurately, by Kayser and Runge, 7 who employed the chloride and carbonate, and measured 162 lines. Barium compounds are gradually dissociated in a hot Bunsen flame, and all exhibit the band-spectrum of the oxide, together with line A. = 5 535. 69 of the metal. Immediately on their introduc- tion the haloid derivatives produce their own peculiar fugitive spectra; these can always be obtained with certainty if a wire holding ammonium chloride is placed in the flame below the specimen of barium salt under examination. For prolonged experiments hydrogen chloride, hydrogen bromide, or iodine vapor must be introduced into the flame. The flame-spectra of these compounds have been studied by Mitscherlich 8 and Lecoq de Boisbaudran. 1 A. B. A. 1861. * P. T. (1864) 154, 139. 3 N. A. S. U. (1868) [3] 6. 4 Spectres lumineux (Paris, 1874). 6 P. T. 163, 369; 164. 806. * Ibid. (.1883). 174, 216. 7 A. B. A. 1891. 8 P. A. (1862) 116, 419; (1863) 121, 459. For the flame-spectrum see also Bunsen and Kirchhoff, P. A. 110, 161. Bunsen, P. A. (1875) 155, 366. * Only visible in the spark-spectrum. (Thalen.) ! Only visible in the spark-spectrum. (Hartley and Adeney.) SPECTRA OF THE ELEM. 103 Arc and 6497.07 5853.91 55I9-37 4700.64 4523.48 3995-92 3599-60 277I-5I spark 6141.93 5826.50 5424-82 4691.74 4506.11 3993.60 3544-94 2634.91 spectra : 6111.01 5805.86 5267.20 4673.69 4432.13 3938.09 3525.23 2347.67 6063. 5800. 4934. 4579- 4402. 3935- 3501. 2335- 33 48 24 84 75 87 29 33 6019.69 5777.84 4903." 4574.08 4350.49 3910.04 3357-00 2304.32 5971- 5680. 4900. 4554. 4283. 3891. 3071. 94 34 13 21 27 97 7i 5907-88 5535.69 4726.63 4525.19 4130.88 3611.17 2785.22 Flame-spectra: Barium bromide 5411 5359 (5305 5250) 5207 5150 Barium chloride 5314 5243 (5206 5172) 5*37 Barium iodide 5608 5377 Barium oxide 6450 6298 (6240 6179 6109 6032) (5939 5868) 5825 (5769 5720 5648) 5535.69* 5493 5347 5216 5090 4874- BERYLLIUM. The whole of the spectrum of this element has not hitherto been thoroughly investigated; some of the visible lines in the spark-spectrum have been measured by Thalen' and Kirch- hoff, 8 and Hartley* has observed others in the ultra-violet. Cornu * mentions two lines in the arc-spectrum, and Crookes * states that, when caused to fluoresce in a vacuum, a continu- ous blue spectrum is produced. Rowland and Tatnall 6 have recently examined the arc-spectrum between \ = 21004600; the lines are comparatively feeble. Arc and spark spectra: 4572.869! 4489-4! - 3905- 2f 3322. 3f (3321.486 3321. 2i8> (3131.200 3130.556) (2651.042 2650.414) 2649.8! (2494.960 2494.532) 2493. 6f 2478.1! 2348.698 1 N. A. S.'U. (1868) [3] 6. - A. B. A. 1861. * J. Chem. Soc. 43, 316. 4 Spectre normal du soleil (Paris, 1881). 5 A. c. p. [5] 23, 555. See also Lockyer, P. Wien. Ber.xjY] 82, 425. 6 Astrophys. Jour. (1895) 1, 16; 2, 185. R. S. 27, 280. Ciamiciaru * Due to the metal itself. f Spark spectrum. 104 SPECTRU'M ANALYSIS. BISMUTH. The spark spectrum is obtained by the use of bismuth electrodes, and has been measured by Huggins, 1 Thalen, 2 and Hartley and Adeney; 3 the arc-spectrum by Liveing and Dewar, 4 and recently, commencing at 6i8/*ju, by Kayser and Runge. 5 The spark-spectrum exhibits many lines that are absent from that of the arc. Bismuth salts, moistened with hydrochloric acid, produce in the Bunsen flame a fugitive band-spectrum of the oxide. The spectra of the compounds themselves are obtained by volatilizing them in a hydrogen flame. They have been drawn by Mitscherlich. 6 Arc and spark spectra: 6493.8* 5271.1* 4260.1* 3654.7! 2938.41 2627.99 2203.2 2110.35 6130.2* 5209.0* (4122.01 ' 3614. 6! 2898.08 2524.58 2189.70 2061.77 6057- 5144. 4121. 3596. 2809. 2515- 2187. 7* 0* 69) 26 74 72 4t 5862. 5124. 4079. 3ii5. 2784. 2489. 2176. 6* 5* 7t 2f 4t 5 70 58I7.I* 4993-9* 3864. 4f 3067.81 2780.57 2400.98 2157.03 5717.6* 4722.72 3793-3t 3024.75 2766.7! 2276.64 2152.98 5552 456o 3757 2993 2730 2230 2134 44 .9* ,6f .46 .61 .70 .38 5451-0* 4302.6* 3695.7f 2989.15 2696.84 2228.31 2133.72 BISMUTH OXIDE. Flame-spectrum. The bands are measured from the red. 6383 6195 6040 5874 5718 5583 5445 5329 5221 BORON. According to Kayser and Runge 7 the arc-spectrum of this element consists of only two lines, which, together with a 1 P. T. (1864) p. 139. 2 N. A. S. U. (1868) [3] 6. 3 P. T. (1884) 175, 130. 4 Ibid. (1883) 174, 222. P. R. S. 29, 398. 5 A. B. A. 1893. P. A. (1863) 121, 459. See also Angstrom, Ibid. (1855) 94, 141. Mas- cart, Ann. de 1'Ecole normale (1866), 4, 7. E. Becquerel, C. r. 96, 1218 ; 97, 72. 'A. B. A. 1892. * Visible only in the spark-spectrum. (Thalen.) ! Visible only in the spark-spectrum. (Hartley and Adeney.) SPECTRA OF THE ELEMENTS. 10$ third, have also been observed by Hartley ' in the spark- spectrum; Eder and Valenta 2 found, however, fourteen additional lines, the majority of which are double, and con- firmed the presence of four that had been detected by Ciamician. 3 Rowland and Tatnall 4 have recently photo- graphed the arc-spectrum between A, = 2 100-4400. Onl) r the double line could be detected; the numerous bands are probably due to some compound, such as boric anhydride. Boric acid and its salts produce a characteristic band spectrum in the Bunsen flame. Arc and spark spectra: 3450.8* (2497.821 2496.867) 2267.0! 2266.4! BORIC ACID. 5 The wave-length is measured at the middle of the bands. Flame-spectrum: 6398 6211 6032 5808 5481 5440 5193 4912 4722 4530 BROMINE. Bromine vapor gives a line-spectrum with the electric spark, 8 but the measurements of it are only approximate. Its absorption-spectrum at the ordinary temperature has been accurately investigated by Hasselberg; 7 when a high disper- 1 P. R. S. 35, 301. 2 Denkschr. d. Wien. Akad. (1893) 60, 307. 3 Sitzber. d. Akad. d. Wiss. zu Wien. [2] 82, 425. See also Troost and Hautefeuille, C. r. (1871) 73, 620. Salet, A. c. p. (1873) [4] 28, 59. 4 Astrophys. Jour. (1895) 1, 16. 5 Lecoq de Boisbaudran, Spectres lumineux (Paris, 1874). Thalen, Up- sal. Universit. Arsskrift. 1866. Also Salet and Eder, and Valenta, as above. 6 Salet, Spectroscopie (Paris, 1888). A. c. p. [4] 28, 26. Plucker, P. A. 105, 527; 107, 87. Plucker and Hittorf. P. T. 155, i. Ciamician, Wien. Ber. 76 [2], 499; 77 [2], 839; 78 [2], 867. 7 K. Svensk. Akad. Handlingar. (1891) 24, No. 3. Mem. de 1'acad, de St. Petersb. (1878) 26, No. 4. See also Daniell and Miller, P. A. 28, 386. Roscoe and Thorpe, P. T. 167, 209. Moser, P. A. 160, 188. * Visible only in the spark-spectrum. (Hartley.) \ Visible only in the spark-spectrum. (Eder and Valenta.) io6 SPECTRUM ANALYSIS. sion is employed it is seen to consist of a large number of fine lines grouped into bands. The spectrum obtained with a continuous discharge differs from that produced when a condenser is included in the circuit. 1 Spark-spectrum of bromine vapor: 7000* 6780 * 6630* 6583 6559 6546 6353 6148 5876 5830 5723 55QO (5509 5497 5491) (5450 5423) (5327 5305 5240 5184 5166) 5060 4930 (4816 4788) 4705 4677 4618 4366 3980 Absorption-spectrum : Group 6162 6142 : 6158.09 6I55.45 6154.19 6150.57 6148 .15 Group 6142 6122 : 6125.12 '6124.01 Group 6122 6103 : 6117.61 b (6116.47 6115.67) 6108 49 Group 6103 6079 : 6083.20 6082.50 6080. 36 6079.78 Group 6079 6066 : 6077-85 6069.61 Group 6066 6042 : 6064. 50 6055-95 6055-76 6055 22 6054 -78 6054.25 Group 6042 6003 : 6021.02 6020. i 5 6018.40 6017.73 b (6017 .18 6016.56) 6013.46 (6011,28 6OII.O2) 6008 .26 6008.03 Group 60035977 : 5982.65 5982.34 5981.55 5981.30 5980.25 Group 59775949: 5966.29 5964-97 5964.35 5961.44 596i. 06 5960.54 5960.16 5959 22 5958.32 Group 59495935 : 5942-73 Group 59355896: 5929-33 5924.62 5924.23 5924.00 5906 13 5904-36 5903-69 5898.45 Group 58965862 : 5869.91 Group 58625832: 5854.95 5846.34 5844.78 5843-44 5838. 81 5838.16 5837.59 5836.95 5836.41 5832. 43 Group 58325807: 5828.60 5812.41 5812.18 5809.57 5808. 87 5808.15 5807.95 Group 58075791: 5802.82 5802.26 5800.93 579S.I5 5796 Si 5795.64 5795-44 Group 57915763: 5782.46 5782.01 5777-39 5774-75 5773- 90 5773-02 577L29 5765.71 5765.03 5764. 42 Group 57635742: 5760.33 5748.57 5747-55 5744.46 5742. 54 1 Trowbridge and Richards, Amer. Jour. Sci. (1897) [4] 3, 117. P M 43, 135. * Inaccurate. SPECTRA OF THE ELEMENTS. ID/ Group 57425712: 5739.81 5730.97 5738. 5729. 9i 5738. 69 b(5727. ii 63 5737- 5726. 25 98) 5736. 5726. 49 5725- 79 5725- 54 5715. 38 5712.44 5712. 18 Group 57125688: 57H. 38 5702. 53 5701. 83 5698. 62 5697- 29 5693- 89 Group 56885659: 5678.02 5676. 93 5667. 97 5667.20 5666. 46 (5664. 31 5664. 06) 5663. 52 5659- 37 Group 5659 5616 : 5657.45 5656. 93 5647.46 5646. 42 5634. 43 5617- 61 Group 56165587: 5603.27 5602. 44 5598. 70 5596. 17 5595- 69 5595. 17 5592.68 Group 55875555: 5573 -63 5572.93 5572.73 b (5570.47 5569- 90 ';!;'v ' '"*."' 5569- 56) (5566. 75 5566. 33) 5565. 97 5563- 90 o?: -^^\ 5563- oo 556i. 35 5550. 44 556o. 16 5559- 86 5557. 42 5556.93 5556. 39 5556. 03 5555- 86 Group 55555528: 5542. 08 5530. 65 Group 55285502 : (5506. 88 5506. 36) 5504. 72 5503. 62 Group 55025477 : 5500. 58 5498. 60 5496. 41 (5491.98 5491- 36) (5485. 19 5484- 93) (5482. 70 5482. 53) 5482. 17 5479- 32 5478. 96 Group 54775456: 5468. 94 5467. 24 (5466. 84 5466. 71) 5464- 96 546i. 81 5460.74 5460. 19 5457- 15 5456. 89 Group 54565430: 5455- 62 5442. 86 5439- 75 (5436.98 5436. 74) 5432. 25 Group 54305406 : 5420.61 5418. 23 54I7. 50 54I7. 03 54I5- 96 5414. 76 5413. 91 5410.93 5408. 9 1 Group 54065372 : 5390. 90 5389. 92 5377-93 5375- 92 5372. 3 5372. ,11 Group 53725354: 5356, 49 5353-So Group 53545333: (5348. 06 5347- 87) 5342. 38 5341. 82 5340. 92 (5338. 31 5338. 07) 5333- ii Group 53335317: (5329. 60 5329. 30) (5319.77 53I9- 56) Group 53175289: (5290 47 5290.21) Group 52625243 : 5256, .32 5255.79 5251- 87 5249- 73 5248. So 5247- 61 5245. 33 Group 52435215: 5241. 88 5241. 72 5234- 00 5215- 92 Group 51845159: 5184. 57 5184. 29 5183.60 5178. 56 5174- 91 5168. 26 5167.41 5163.35 5160. 54 CADMIUM. The references given below show how numerous have been the investigations of the cadmium-spectrum; the latest are those of Kayser and Runge, 1 who occasionally employed the 1 A. B. A. 1891. IO8 SPECTRUM ANALYSIS. chloride, but usually the metal. The arc-spectrum differs considerably from that of the spark; the latter exhibits a pair of lines of the highest intensity of A = 53/8.8 and 5338.3, which are absent from the former, and the same applies to 90 lines of 1 inferior brightness between X = 4215 an d 21 11 which have been measured by Hartley and Adeney. 1 Cad- mium chloride and bromide are dissociated in the Bunsen flame, and exhibit the lines A = 5086, 4800, and 4678. Arc and spark spectra : 6467.3* 6438.8* 5378.8* 5338.3* 5154.85 5086.06f 4800.09f 4678. 37f 4662.69 4413.23 361304 3610.66 3467.76 3466.33 3403.74 3261.17 (3252.63 3133.29 3081.03) 2980.75 2880.88 2763.99 2639.63 2601.99 2573.12 2329.35 2312.95 228810 2267.53 2239.93 2194.67 2144.45 CESIUM. 2 Bunsen and Kirchhoff discovered caesium in 1861 by means of spectrum analysis. Its salts are all dissociated in the Bunsen flame, and exhibit the lines of the metal; the more prominent ones are A = 4555 and 4593 in the blue, and A = 6010 in the orange. Arc, spark, and flame spectra: 6973.9 6723.6 6213.4 6010.6 5845.1 5664.0 5635.1 4593-34 4555-44 3888.83 3876.73 3617.08 3611.84 1 P. T. (1883) 175, 98. See also Thalen, N. A. S. U. (1868) [3] 6. Kirch- hoff, A. B. A. 1861. Mascart, Annales de 1'Ecole hormale (1866), 15. Lockyer, P. T. (1873) 163, 369. Cornu. Journ. de Phys. (1881) 10, 425. Liveing and Dewar, P. R. S. (1879) 29, 482. P. T. (1888) 179, 231. Ames, P. M. (1890) [5] 30, 33. Bell, Am. Journal of Sciences, June, 1886. 2 Kayser and Runge, A. B. A. 1890. Bunsen, P. A. 119, i; 155, 366. Johnson and Allen, P. M. [4] 25, 199. Thalen, N. A. S. U. (1868) [3] 6. Lecoq de Boisbaudran, Spectres lumineux (Paris, 1874). Lockyer, P. R. S. {1878) 27, 280. Liveing and Dewar, Ibid. (1879) 28, 352. * Only visible in the spark spectrum. (Thalen.) f Visible also in the flame spectrum of the chloride and bromide. (Lecoq de Boisbaudran.) SPECTRA OF THE ELEMENTS. 109 CALCIUM. The line-spectrum of calcium has been measured by Kirchhoff, 1 Muggins," Thalen, 3 Lecoq de Boisbaudran, 4 Lockyer, 5 Cornu, 6 Liveing and Dewar, 7 and more recently by Kayser and Runge, 8 who employed the electric arc and calcium chloride. The faint bands which occasionally appear in the yellow and red when the arc is used are considered by Lecoq de Boisbaudran to be due to oxide. Many of the calcium lines are readily produced, and are therefore always visible when carbon electrodes are employed. H. Becquerel* observed bands from A = 8880-8830 and from A = 8760-8580 in the infra-red. The haloid compounds have been investi- gated by Bunsen, Mitscherlich, 10 and Lecoq de Boisbaudran; 1 * in the Bunsert flame some bands peculiar to each compound are visible, together with those of the oxide and the blue line, A = 4226.91, of the metal. The oxide bands are also- produced if the flame is charged with hydrogen chloride,, hydrogen bromide, hydrogen iodide, or hydrogen fluoride. Arc and spark spectra: 6499.85 6462.75 6439.36 6169.87 6162.46 6122.46 6102.99 5867.94 5857.77 5603.06 5601.51 5598.68 5594.64 5590.30 5588.98 5582.16 5513-07 5349.66 5270.45 5265.79 5264.46 5262.48 5261.93 5189.05 5041-93 4878.34 4586.12 4581.66 4578.82 4527-17 4456.08 4454.97 4435.86 4435.13 4425.61 1 A. B. A. 1861. 2 P. T. (1864) 154, 139. 3 N. A. S. U. (1868) [3] 6. Spectre du fer (Upsala, 1884). 4 Spectres lumineux (Paris, 1874). 5 P. T. (1874) 164, 809. C. r. 82, 660. 6 Spectre norm, du soleil (Paris, 1881). I P. T. (1882) 174, 187. P. R. S. 28, 367, 475; 29, 398. * A. B. A. 1891. See also Mascart, Ann. de 1'ecole normale (1866), 15. Angstrom, Recherches sur le spectre solaire (Upsal, 1868). Eder and Va- lenta, Phot. Corresp. 1893, p. 59. Rydberg, W. A. (1894) 52, 119. 9 C. r. 94, 1218; 97, 72. 10 P. A. 121, 459 II Spectres lumineux (Paris, 1874). I I O SPECTR UM ANAL YSIS. 4355- 4i 43i8, ,80 4307.91 4302.68 4299. 14 4289.51 4283.16 4226 91* 3973. ,89 3968.63 3957.23 3933. 83 3644.45 3630.82 3624. 15 3487. ,76 3361.92 3350.22 3344- 49 3I79.45 3158.98 239. 66 2275 .60 CALCIUM BROMIDE. Flame-spectrum : 6267 6243 6103 CALCIUM CHLORIDE. Flame-spectrum: 6266 (6203 6182) (6069 6045) 5934 5817 (5544 55i8)f CALCIUM FLUORIDE. Flame-spectrum: 6061 6027 5329 5302 CALCIUM OXIDE. Flame-specUupi : ' \ * 0221 5996 (5544 5518) CARBON. This element htfs been more frequently investigated than any other; its spectra are extremely complex, but it is now generally acknowledged to exhibit two, a line and a band spectrum. The former is produced by the passage of sparks from a Leyden jar through carbon dioxide or carbon mon- oxide; its visible portion, between wave-length 3920-2266.5, has been measured by Angstrom and Thalen, 1 and the ultra- violet part by Liveing and Dewar. a The band-spectrum has also been termed the " flame-spectrum," or " Swan's spec- trum" ; it was first observed by Wollaston 3 in 1802, and then 1 N. A. S. U. (1875) [3] 9- 2 P. R. S. (1880) 30, 152, 494; (1882) 33, 403; (1883) 34, 123. P. T. (1882) 174, 187. Also Eder and Valenta, Denkschr. Wiener Akad. (1893)60. 8 P. T. 1802. * Also visible in the Bunsen flame. f Probably due to the oxide. SPECTRA OF THE ELEMENTS. Ill investigated by Swan 1 from 1850 onwards. In common with Angstrom, Thalen, 2 and Liveing and Dewar, 3 he ascribed it to hydrocarbons, but the last workers, together with Attfield, 4 Morren, 5 and Dibbits, 8 subsequently recognized that it was due to carbon, since it is produced by the combustion of pure cyanogen in dry oxygen. This band-spectrum consists of five complex bands, with the following wave-lengths according to Angstiom and Thalen, and Watts: 7 i or orange band. y^ilow^anV 3 or green band. 4 or blue band. 5 or indigo band 6187-5954 5633-5425 5164-5082 ! 4736-4677 4331-4232 The three medial bands have been recently measured by Fievez, 8 and the green one by Kayser and Runge; 9 for the others there are only the old observations of Watts, Angstrom and Thalen, and Piazzi- Smyth 10 available. In addition to the above bands others are sometimes observed in the arc; they occur in the blue, violet, and ultra-violet, and have the following wave-lengths: I Band. II Band. Ill Band., IV Band. V Band. 4600-4500 4290-4150 3884-3856 3590-3550 3370-3350 Their existence in the arc is doubted by Kayser and Runge; Liveing and Dewar have ascribed them to cyanogen, whilst Lockyer, 11 H. W. Vogel, and others regard them as a 1 P. T. E. (1857)21,411. 2 N. A. S. U. (1875) 9. 3 P. R. S. (1880) 30, 152, 4945(1882) 33, 403; (1883) 34, 123. P. T. (1882) 174, 187. 4 Ibid. (1862) 152, 221. P. M. (1875) 49, 106. B 'A',c. p. (1865) [4] 4,305. B P. A. (1864) 122, 497. I P. M. (1869) [4] 38, 249; 45, 12; (1874) 48, 369, 456; (1875) 49, 104. 8 Mem. de 1'Acad. roy. de Belgique (1885), 47. 9 A. B. A. 1889. 10 Astr. Obs. Edinb. (1871) 13, 58. P. M. (1875) [4] 49, 24; (1879) [s] 8, 107. P. T. E. 30, 93. II P. R. S. (1878) 28, 308; (1880) 30, 335. See also PlUcker, P. A, (1858) 105, 77; (1859) 107, 533, and with Hittorf, P. T. 155, i. Jahresber. (1864) p. no. Van der Willigen, P. A. (1859) 107, 473. Huggins, P. T. (1868) 112 SPECTRUM ANALYSIS. second band-spectrum of carbon produced only at high tem- peratures. Kayser at first shared this view, but experiments made in conjunction with Runge led to a different conclusion. A strong current of carbon dioxide was directed on to the arc, whereupon the cyanogen bands became fainter and dis- appeared. In order to prove that this was not due to a lowering of the temperature a still stronger current of air was substituted for the carbon dioxide: the bands immediately increased in brightness in consequence of the additional supply of nitrogen. The view that the cyanogen bands are essentially due to carbon is supported by their occurrence in comets, and in the solar spectrum; this last fact was long doubted, but was established by Rowland. That the spec- trum of a compound which is dissociated at 1000 should be visible in the solar spectrum appears somewhat paradoxical, but Kayser and Runge have pointed out that the carbon molecule, as is shown by its varying specific heat, is not a constant quantity, and the "cyanogen bands" maybe the spectrum of an unknown compound of carbon and nitrogen which is capable of existence at very high temperatures. The cyanogen bands have been measured by Liveing and Dewar, and the third, fourth, and fifth ones also by Kayser and Runge. The carbon bands all have their brighter edges directed towards the red end of the spectrum; each possesses a number of edges, varying from three to seven, which become weaker towards the violet. The lines extend from the first edge of one band to the beginning of the next, so that no portion of the spectrum from 620^ to 34OyUyu is free from carbon lines, the total number of which is at least 10,000. Metallic spectra obtained by means of the arc and carbon poles always exhibit 158, 558. Lielegg, Wien. Ber. (1868) 52. 593. Troost and Hautefeuille, C. r. (1871) 73, 620. Wiillner, P. A. (1872) 144, 481. Salct, A. c. p. (1873) [4] 28, 60. Lecoq de Boisbaudran, Spectres lumineux (Paris, 1874). Wesen- donck, Inaug-Diss. (Berlin, 1881). Hartley, B. A. R. 1883. Eder, Denkschr. Wiener Akad. (1890) 57. SPECTRA OF THE ELEMENTS. 11$ the metallic lines superposed by the carbon bands, hence a good knowledge of the latter is very desirable. Line-spectrum: 6584.0 6578.5 5695.1 5661.9 5647.5 4266.6 3920.0 2837.2 2836.3 2746.5 2509.0 2478.3 2296.5 Band-spectrum: I. Orange Band 6187 5954: 6188.2 6119.9 6057.3 5145.0 2511.9 6001.8 5954-5 II. Green-yellow Band: i. Edge: 5635.3 5634-3 5633.9 5633.4 5632.9 5632.2 5631-6 5630.9 5630.1 5629.3 5628.5 5627.6 5626.7 5625.8 5624.4 5623.1 5620.4 5619.6 5618.8 5617.8 5615.6 5613-8 5611.8 5610.1 5608.5 5606.1 5603.5 5600.8 5598.0 5595-3 5592-5 5589-7 5588.1 5587.4 5585.5 2. Edge: 5584.4 5582.1 5578.8 5578.5 5578.2 5577-1 5575-5 5573-2 5572-8 5572-4 557L I 5570.1 5569.5 5568.3 5567.2 5566.7 5565.0 5564-3 5563.9 5563.5 5561.4 5560.5 556o.o 5558.1 5556.3 5556.0 5555-7 5554-1 5552-9 5552.6 5540.9 (3- Edge) ; . III. Green Band: I. Edge: 5165.4 5165.2 5164.9 5164.6 5164-5 5163-7 51^3.2 5162.7 5162.5 5162.0 5161.3 5160.5 5160.0 5158.7 5157.8 5157.3 5156.2 5155.3 5I54.5 5154-4 5I53.4 5153.0 5152-6 5152.0 5150.8 5149.2 5147-8 5146.2 5144.7 5143.0 5I4I-3 5139-4 5137.7 5135.7 5133.8 5I3L7 5129-7 5129.4 2. Edge: 5129.4 5129-3 5129.0 5128.8 5127.8 5127-4 5126.2 5125.4 5124-9 5123-9 5122.9 5121.8 5121.6 5II9.3- 5118.2 5116.8 5H5.9 5H4.4 5113.2 5111.8 5110.8 5110.2 5109.2 5108.0 5106.5 5105.5 5103.9 5103.5 5102.6 5101.0 5099.9 5098.2 5096.9 5095.3 5094.2 5092.4 5091.6 5091-0 5089.3 5087.6 5087.1 5086.4 5084.9 5083.1 5081.9 5080. i 5078.5 5076.8 5071.9 5070. i 5068.8 5066.9 5066. i 5065.1 5063.7 5058.1 5056.3 5054.8 5052.8 5049.7 5045.4 5041-5 5039.9 5037.9 (5033.9 5033-7) 5032.2 5030.1 5026.0 5024.1 5022.1 5017.9 50I3-9 5009.6 5005.6 5001.1 4997.0 4992.5 4988.3 4983.7 4979-4 4974.6 4970.3 4965.4 4961.0 4956,1 4951.6 4946. 5 4886.2 4854.2 4849.0 4838.0 4809.7 4804.4 4798.4 IV. Blue Band: i. Edge: 4737-2 4736.4 4735-9 4735-5 4735-1 4734-6 4734-1 4733-6 4733-0 4732-4 4732-0 4731-0 4730.0 4729.4 SPECTRUM ANALYSIS. 4728.4 4727-1 4726.2 4725.1 4724.0 4723.6 4722.3 4721.2 4719-9 4718.8 4717-4 4716.8 4715.4 2. Edge: 4715.2 4714-1 47I3.3 47II-7 47io.5 4708.6 4707.2 4706.9 4705.9 4704.0 4702.6 4702.1 4699-9 4699.4 4698.4 4697.6 3- Edge: 4697.2 4695-6 4695.3 4694.2 4693.9 4692.8 4691.2 4690.7 4689.5 4688.3 4687.4 4686.2 4685.0 4. Edge: 4684.7 4683.2 4682.5 4682.0 4681.0 4679-8 4679.5 4677.6 4675.8 4675.0 4674.1 4673-8 4673-1 4672.8 4671.8 4671.4 4671.0 4670.4 4669.8 4669.2 4668.6 4667.7 4666.9 4666.5 4665.6 4664.4 4664. i 4663.0 4661.3 4659-9 4658.9 4657.9 4657.1 4656.8 4654-6 4653-7 4652.3 4652.0 V. Band: i. Edge 4382.0 2. Edge 4371.4 3. Edge 4365.1 5182.5 5I79.5 4823.55 4821. 3 b 4394. 7 b r (A. andTh.) CARBON MONOXIDE. 1 6079 5609 5198 b r 5187.5 5184.5 5176.0 5173.5 5170 5167.2 4834.5 b' 4819. i b 4817.0 b (4789.8 b 4786. 5 b) 4509.8 b* 3493-3 b r 3307. 5 b>- 3134.6^ 2976.3 b^ 3832.0 b r 2792.7 b r 2665.1^ 2599. ob* 2510.8^ 2435. ob* 2425. ob* 2404.7^ 2394.0 b v 2381. 5 b 2364.8 b* 2337. 7 b v 2311. 4b* 2286. 2 b v 2220 b v 2215. 3 b v 2188. i b" 2172.3 b* 2161. 6 b* 2149.9 b* 2127. 8 b r 2112. 7 b v 2089. 3 b r 2066.8 b' <(Deslandres). The bands marked b r are sharply defined towards the red and shade off towards the violet ; those marked b" are sharply defined towards the violet and shade off towards the red. I. Band: 4601 4575 CYANOGEN." 4551 4533 45i6 4506 4501 II. Band: i. Edge: 4216.17 4215-67 4215-31 4215-04 4214.76 4214.08 4213.71 4213.29 4212.85 4212.39 4211.90 4211.37 4211.03 4210.82 4210.23 4209.62 4208.98 4208.29 1 Angstrom and Thalen, N. A. S. U. (1875) 9. Deslandres, A. c. p. 5525.27 5523-56 5489.90 5484.22 5483.57 5477-13 5454-79 5444-Si 5407.75 5381.99 5369.79 1 P. T. 1881. Part 3. 2 Spectre normal du soleil (Paris, 1881). 3 Svenska, Vetensk. Akad. Forhandl. (1896) 28, No. 6. From D A. 3450. Astrophys. Jour. (1896) 3, 288; 4, 343; (1897) 5, 38. 4 P. T. (1888) 179, 231. From A. 3450-2244. 5 Ber. 11, 916. Monatsber. Berl. Akad. 1878, p. 415. 6 P. R. S. 31, 51. Ber. 14, 503. 7 J. Chem. Soc. 1889, p. 14. Ber. 24, 619. 8 Zeitschr. Anal. Chem. 18, 38. See also Etard, C. r. (1895) 120, 1057. * Visible only in the spark-spectrum, 120 SPECTRUM ANALYSIS. 5362.97 5359.41 5353.69 5352.22 5343.58 5342.86 534L53 5331.65 5325.44 5316.96 5312.84' 5301.24 5280.85 5276.38 5268.72 5266.71 5266.51 5257.81 5248.12 5235.37 5230.38 5212.87 5176.27 5156.53 5154.26 5146.96 5133.65 5126.37 5125.88 5123-01 5H3.4I 5109.08 5095.18 4988.15 4980.15 4972.16 4966.77 4928.48 4904.37 4899-72 4882.90 4868.05 4843.61 4840.42 (4814.16 4813.67) 4796.00 4793.03 4785.26 ' 4781.62 4780.14 4778.42 4776.49 4771.27 4768.26 4767.33 4754-59 4749.80 4737-95 4735.04 4728.14 4718.67 4698.60 4693.37 4682.53 4663.58 4657.56 4644.48 4629.47 4625.88 4623.15 4597-02 4594-75 4581.76 4580.32 457o.i8 4566.77 4565.74 4549.80 4546.14 4543.99 4534.18 4531.14 4517.28 45I4.33 4494.92 4484.07 4483.70 4478.45 4471.70 4469.72 4467.04 4445.88 4421.48 4417.55 4392-02 4391.70 4380.25 4375-09 4373-77 4371.27 4339.76 4331.38 4303.36 4285.93 4252.47 4234.18 4190.87 4162.33 4158.58 4121.47 4118.92 4110.69 4086.47 4082.76 4077.55 4076.28 4068.72 4066.52 4058.75 4058.36 4053.08 4045.53 4035.73 4027.21 4021.05 3998.04 3995.45 3979-65 3978.8o 3974-87 3973.29 3969-25 3958.06 3953.05 3945-47 3941.87 3941.01 3936 12 3922.88 3917.26 3910.08 3906.42 3895.12 3894.21 3884-76 3882.04 3876.99 3874-10 3873.25 3861.29 3851.09 3845.95 3842.20 3816.58 3816.46 3755-59 3750.06 3745-61 3736.05 3734.30 3733-62 3732.52 3730.61 3708.96 3702.40 3693.65 3693.27 3684.62 3683.18 3676.69 3652.68 3649.47 3647.82 3643.34 3641-95 3639-63 3634.86 3633.00 3631-55 3627.96 3625.13 3611.89 3605.50 3595-00 3587.30 3585-28 3584.92 3575.48 3575.06 3569.48 3565-08 3563-04 3562.22 3561.01 3558.90 (3553-12 3552.85) 3550.72 3548.6o 3543.40 3533-49 3529.92 3529-17 3526.96 3523.85 3523-57 3521.70 3520.20 3518.49 3513-62 3512.78 3510.53 3509-98 3506.44 3502.76 3502.41 3496.83 3495.82 3491.46 3490.89 3489.54 3485.49 3483.55 3474-15 347L52 3466.0 3462.9 3453-6 3449-3 3443-7 (3434.0 3433.5) 3432.0 (34I3-7 34I3.4 3410-3 3406.0 3396.5 3389-3 3388.8 3381.7 3368.3 3356.4 3336.1) 3284.9 3158.6 3154-6 3147.0 3I39.9 3137.4 3121.5 3086.7 3082.5 3072.4 3061.9 3049-0 3044.0 2989-5 2986.9 2954.5* 2942.9 2824.9 2648.8 2580.2 2564.0 2541.9 2540.6 2533-8* 2532.1 2530.0 2528.5 2524.9* 2524.6 2521.1 2519.7* 2510.9 2506.2 2497-5 2490. 2 2464. i 2436.9 2432.4 2420.7 2417.6 2411.6 2407.5 2397-3 2388.8 2378.5 2363-7 2353-4 23I3-9 23H.5 2307.8 2293.4 2286.1 2266.6* 2260.1* 2256.8* 2245.2* * Visible only in the spark-spectrum^ SPECTRA OF THE ELEMENTS. 121 COPPER. The spark-spectrum of copper, in the visible field, has been measured by Kirchhoff, 1 and Thal6n, 2 and, as far as wave- length 4275, by Lecoq de Boisbaudran ; 3 in the ultra-violet Hartley and Adeney 4 have measured the portion between A = 3999 and 2103, and Trowbridge and Sabine 5 that between A = 2369 and 1944. Liveing and Dewar 6 have photographed the arc-spectrum from A = 2294-2135, whilst, more recently, Kayser and Runge 7 have done the same for the region between X. = 6000 and 1944; they measured 304 lines, and obtained the spectrum by substituting, for the carbon poles, rods of copper 1-2 sq. cm. in section. Scarcely any of the copper lines are sharply defined even on one side, so that the spectrum has a peculiar appearance. In the Bunsen flame cupric chloride produces a band-spectrum extending over the whole field, with the exception of the violet; the same spectrum is obtained with the metal if the flame contains hydrogen chloride. The absorption-spectra of copper salts are not characteristic, as the compounds produce total extinction both in the red and the violet. Ewan 8 found that, in aqueous solution, the spectra of the chloride, sul- phate, and nitrate change with progressive dilution tending to become identical; this observation is in agreement with the theory of electrolytic dissociation. C. H. Wolff 9 has sug- gested a spectro-colorimetric method for the determination of A. B. A. 1861. N. A. S. U. (1868) 6. Spectres lumineux (Paris, 1874). P. T. (1883) 175, 63. Proc. Amer. Academy, 1888. P. M. [5] 26, 342. P. R. S. (1879) 24, 402. P. T. (1883) 174, 205. A. B. A. 1892. 8 P. M. 1892, p. 317. Ber. 25, 4950. P. R. S. (1895) 57, 128. 9 Zeitschr. anal. Chem. 18, 38. 122 SPECTRUM ANALYSIS. copper in small quantity, and P. Sabatier * has studied the absorption-spectra of solutions of cupric bromide. Arc and spark spectra: 5782.30 4704.77 4480.59 3602.11 3247.65 2618.46 2303.18 2199.77 5700.39 4674.98 4415.79 3599-20 3126.22 2492.22 2293.92 2178.97 5292.75 4651-31 4378.40 3512.19 3108.64 2441.72 2263.20 2165.20 5220.25 4587.19 3450.47 3063.50 2406.82 2230.16 2104.88 5218.45 4539-98 4259.63 3308.10 3036.17 2392.71 2227.85 2025.14 5153-33 453I-04 4062.94 3290.62 2961.25 2370.5* 2225.77 1943-88 5105.75 4507.62 4022.83 3274.06 2766.50 2369.97 2214.68 CUPRIC CHLORIDE. 2 Flame-spectrum: 6619 b 6268 b v 6151 b 7 ' 6051 b v 5564 b 5507 5440 5386 5306 b 5270 b 5240 5088 b 7 ' 5050 b v 4984 b v 4946 b v 4883 b v 4848 b" 4793 b" 458o fa" 4523 b 4497 b v 4437 b" 4413 b" 4354 b w 4332 b" 4282 b v 4261 b v DIDYMIUM. 3 The metals of the rare earths have been frequently inves- tigated during the past few years, and new substances have been discovered in bodies which were formerly believed to be elementary. In 1885 didymium was resolved into neodymium and praseodymium, and samarium, which was discovered in 1879, was also shown to be present. Other members of the cerium and yttrium groups have likewise been decomposed into its elements. At present the chemical properties of these 1 C. r. 118, 1042, 1144. Ber. 27, 489, 490. See also Glan. W. A. (1878) 3, 65. Kriiss. Colorimetrie (Hamburg, 1891). 2 Lecoq de Boisbaudran, Spectres lumineux (Paris, 1874). Diacon, A. c. p. (1865) [4] 6, 5. A. Mitscherlich, P. A. 116, 499; (1863) 121, 459. 3 Thalen, Om Spectra tillhoranda Yttrium, Erbium, Didym och Lan- than (Stockholm, 1878). Ofersigt K. Vetenskaps Akad. Forhandl. (1883) 40, No. 7. Gladstone, J. Chem. Soc. 10, 219. Bunsen, P. A. 155, 366. Kirch- hoff, A. B. A. 1861. Delafontaine, P. A. 124, 635. Lockyer, P. T. 1881. * Visible only in the spark-spectrum. (Hartley and Adeney.) SPECTRA OF THE ELEMENTS. 12$ substances have not been sufficiently studied to render their recognition as elements absolutely free from doubt. Kriiss and Nilson, as the result of their investigation of the absorp- tion-spectra, consider that didymium, erbium, holmium, samarium, and thulium are composed of more than twenty elements; their conclusion is based on the assumption that each element has a characteristic maximum of absorption, but Schottlander's extensive investigations show that this is falla- cious. Bailey has also raised objections to Kruss and Nil- son's conclusions. At present the results of the spectroscopic work on the rare earths is so uncertain, that the data given in this book usually refer to the " old " elements. Bahr and Bunsen found that didymium oxide, like erbium oxide, when heated in the Bunsen flame gives a continuous spectrum, and also characteristic bright lines which are almost coincident with the absorption-lines exhibited by solutions of the salts, or by glass which contains the metal; this is no exception to the rule that solids only yield continuous spectra, for Huggins and Reynolds showed that the earths are volatile in the oxy- hydrogen flame. Comparison of the absorption-spectra of didymium chloride, sulphate, and acetate led Bunsen to the conclusion that the bands tend to approach the red as the molecular weight increases. The absorption-spectra of the rare earths in the ultraviolet has been investigated by Soret. Spark-spectrum: 5486.0* 5372.0 5361.5 5319.9 5293.5 5273.5 5249.5 5192.5 5191-5 5130.3 4924.5 4463-2 4452.3 4446.7 4328.1 4303.6 4109.8 4060. 7 * Due possibly to samarium. 124 SPECTRUM ANALYSIS. DIDYMIUM CHLORIDE. Absorption-spectrum : l b (7431-7* 736i.7* 7308.7*) 6895.6* 6793-3* l>]b(5963t 5886f 5824* 5789* 5748* 5720) [/3] b (5313* 5206) (5125.8* 5088*) 4823! 4759 4692f 4441.7 "OLD" DIDYMIUM NITRATE. Absorption-spectrum: positions of minimum of brightness: 7291 6906 6794 6407 6235 6189 5797 5759 5317 5253 5217 5147 5126 4826 4771 4695 4633 4443 4341 4289.6 4173.6 PRASEODYMIUM NITRATE. 7291 6794 5916 5797 5759 5317 5217 5125 4826 4695 4443 ERBIUM. 2 The remarks on didymium apply also to this element. Spark-spectrum : 5827 5763 5344-4 5257 52i8 5189 4952 4899.9 4872.4 4820 4674.9 4606.3 4501.3 1 Lecoq de Boisbaudran, Spectres lumineux (Paris, 1874). C. r. (1887) 105, 276. Bunsen, P. A. 155, 366. Ann. Chem. Pharm. 128, 100; 131, 255. Huggins and Reynolds, P. R. S. 18, 546. Lippich, Sillim. Journ. (1873) [3] 13, 304. Auer v. Welsbach, Sitzungsber. Wien. Akad. (1885) 92. Crookes, C. N. 54, 27. Schuster and Bailey, B. A. R. 1883. H. Becquerel, C. r. 104, 777, 1691; 106, 106. Haitinger, Monatsch. f. Chem. (1891) 12, 362. Soret, C. r. 86, 1062; 88, 422; 91, 378. Kriiss and Nilson, Ber. 20, 2143. Bailey, Ber. (1887) 20, 2769, 3325. Schottlander, Ber. (1892) 25, 569- * Thalen, Om Spectra Yttrium, Erbium, Didym och Lanthan (Stock- holm, 1874). Ofversigt K. Vetensk. Akad. Forhandl. (1881) 40. Bunsen and Bahr, Ann. Chem. Pharm. 137, i. Huggins, P. R. S. 1870. Bunsen, P. A. 155, 366. * Neodymium. f Praseodymium. SPECTRA OF THE ELEMENTS. 12$ ERBIUM CHLORIDE. 1 Absorption-spectrum : 6839 6671 6535 6405 5410 5364 [] 5232 4922 4875 45i6 FLUORINE. There are no accurate measurements of the spectrum of fluorine. By the passage of induction-sparks through silicon fluoride Salet 2 obtained a beautiful blue band-spectrum of the compound; the incission of a Leyden jar produced the spark-spectrum of fluorine. Commencing at Salet's last lines Liveing 3 measured the flame-spectrum. Spark-spectrum: 6922* 6862* 6782* 6401 6231 Flame-spectrum : 6231 6091 6011 5571 5321 GALLIUM. Lecoq de Boisbaudran, 4 who discovered this element, measured its spark-spectrum, and Liveing and Dewar 6 that of the arc. Spark and arc spectra: 4171 4031 GERMANIUM. The spark-spectrum of germanium has been investigated by Kobb, 6 and by Lecoq de Boisbaudran, 7 who calculated its 1 Lecoq de Boisbaudran, Spectres lumineux (Paris, 1874). Bunsen, P. A. 155, 366. Bunsen and Bahr, Ann. Chem. Pharm. 137, i. 2 A. c. p. (1873) 28, 34. 3 Proc. Cambridge Phil. Soc. 3, Pt. 3. See also Mitscherlich, P. A. 121, 476. Seguin, C. r. (1862) 54, 933. 4 C. r. 82, 168. 6 P. R. S. 28, 482. 6 W. A. (1886) 29, 670. 7 C. r. (1886) 102, 1291. Her. 19, 479. * Only approximate. 126 SPECTRUM ANALYSIS. atomic weight from his measurements. Rowland and Tatnall l have recently photographed the arc-spectrum between A, = 2300-4600. Arc and spark spectra: 6337 6021 5893 5256.5 5229.5 5210 5178.5 5135 5131 4814 4743 4685.3 4291-6 4261 4226 4179 3269.628* 3124.945* 3039.198* 2754.698* 2740.535* 2709.734* 2691.446* (2651.709* 2651. 219f) 2592.636* 2417.450* GOLD. Liveing and Dewar 2 measured three lines in the ultra-violet exhibited by the arc-spectrum, and these were the only ones known when Kayser and Runge 3 commenced their investiga- tion of the region between wave-length 6600 and 2280. They usually employed fine gold, but occasionally auric chloride and carbon poles. The visible portion of the spark-spectrum has been measured by KirchhofT, 4 Huggins, 5 Thalen, 6 and G. Kruss, 7 who ascribes the line 5230.47 to platinum, and not to gold. Arc and spark spectra: 6278.37 5957-24 5837-64 5656.00 5250.47 5064.75 4792.79 4488.46 4065.22 3122.88 3033.38 3029.32 2932.33 2905.98 2676.05 2428.06 HELIUM. The yellow Z> 3 -line of the solar spectrum was ascribed, until recently, to a hypothetical element, termed by Frank- 1 Astrophys. Jour. (1895) 1, 149. 2 P. T. (1882) 174, 2219. 3 A. B. A. 1892. 4 Ibid. 1 86 1. 5 P. T. 1864, p. 139. 6 N. A. S. U. [3] 6. 7 Lieb. Ann. (1887) 238, 30. See also Lecoq de Boisbaudran, Spectres lumineux (Paris, 1874). * Arc-spectrum. f Possibly a single reversed line. SPECTRA OF THE ELEMENTS. I2/ land helium; this was isolated by Ramsay 1 in 1895 from cleveite, in which it occurs together with argon; he also obtained it from certain meteorites from Augusta Co., Virginia. Cleve 2 showed that it is present, unaccompanied by argon, in cleveite from Carlshuus in Norway, and he observed the presence in its spectrum of five lines in addition to D a . Deslandres, 3 using a very high dispersion, measured the following lines: 6678 5876.0 5048.4 5016.0 4922.2 4713.35 4471.75 4437.9 4388.4 4I43-9 4120.9 4026.2 3964.0 3888.75 3819.7 3705.4 3613-8 3447-7 3187.9 2945.7 Runge and Paschen, 4 in the course of an investigation of the gases from cleveite, showed that the line at 5876.0 is a double one, and, as the solar helium line had always been regarded as single, doubt was cast on the identity of solar and terrestrial helium. This point was speedily settled by Huggins and Hale, 5 who showed that the solar Z> s -line is also double. Kayser 6 found that a Geissler tube containing what he supposed to be the purest atmospheric argon also showed the Z? 3 -line, ^ us affording proof that helium is present in the atmosphere. Lockyer 7 states that many of the helium lines coincide with some of hitherto unknown origin in the spectra of the chromosphere, and of the white stars of Orion. Under the influence of the silent discharge helium combines with mercury and benzene or carbon bisulphide to form a com- pound resembling that of argon, but it does not combine with mercury alone. 1 C. r. 120, 660, 1049. Ber. 28, 318, 448. N. 52, 224. Ramsay, Collie, and Travers, Jour. Chem. Socy. (1895) 67, 648. 2 C. r. 120, 834. Ber. 28, 373. 3 C. r. (1895) 120, mo, 1331. 4 W. A. Beibl. 19, 634. 5 C. N. 72, 26. 6 Ibid. 72, 99. 1 C. r. 120, 1103. P. R. S. 58, 67. See also Brauner, C. N. 71, 271. Palmieri, Acad. di Napoli Rendic. (1882) 20, 233. 128 SPECTRUM ANALYSIS. Berthelot ' and Runge and Paschen a have observed that the spectrum of cleveite gas consists of six series, two pairs of which are characterized as subseries, whilst two series are principal series. Two spectra are thus differentiated which are ascribed to two constituents of the gas, and which bear a striking similarity to the spectra of the alkalies. Rydberg 3 has confirmed these conclusions, and termed the second con- stituent parhelium. Some confirmation was also afforded to this view by Ramsay and Collie's researches, which resulted in the separation of helium into a lighter and a heavier por- tion; but Ames and Humphreys 4 were unable to detect any difference in their spectra, although they used a spectroscope of high dispersive power. When further separated the heavier portion was found to consist chiefly of argon. HYDROGEN. Two spectra, termed the elementary and compound line- spectra, are exhibited by hydrogen in a Geissler tube; their production depends on the conditions of temperature and pressure. The former has been measured by Angstrom, 5 H. W. Vogel, 6 Lockyer, 7 Huggins, 8 Cornu and Ames; 9 the latter was first investigated by Pliicker and Hittorf, 10 but was ascribed to acetylene by Angstrom, Berthelot and Richard, 11 and Salet. 12 The incorrectness of this view was proved by 1 C. r. (1897) 124, 113. 2 Sitzber. Berl. Akad. (1895) 639, 759. W. A. Beibl. (1895) 19, 884, 885. Astrophys. Jour. (1896) 3, 4. W. A. (1896) 58, 674. Astrophys. Jour. (1896) 4, 91. Astrophys. Jour. (1897) 5, 97. P. A. (1864) 91, 141; 123, 489; (1872) 144, 300. Berl. Monatsber. (1879) 586; (1880) 190. Ber. (1880) 13, 274. P. R. S. 28, 157 ; 30, 31. P. T. 171, 669. P. M. (1890^ [5] 30, 33. 10 P. T. (1865) 155, 21. Pliicker, P. A. 107, 407. 11 C. R. 68, 810. 1035, 1107, 1546. 12 A. c. p. [4] 28, 17. SPECTRA OF THE ELEMENTS. 129 Hasselberg's 1 exact measurements. In the spectrum of C Puppis Pickering 2 found, in addition to dark hydrogen lines and K, two broad lines at A. = 4633 and A = 4688, and a peculiar series of dark lines whose wave-lengths are rhythmi- cally related. These were A 4544, 4201, 4027, 3925, 3859, 3816, 3783. It was first thought that they represented some new element not yet found on the earth or in the stars, but they are very probably due to hydrogen, produced under conditions of luminosity hitherto unknown. By applying Balmer's formula, Pickering found that the new lines form a harmonic series. This conclusion was confirmed by Kayser,* who pointed out that hydrogen had been the only element, having harmonically related lines, which had possessed only a single series of such lines. Kayser and Runge had pre- viously found that two of the series of lines of an element end at nearly the same place. On examining the oscillation frequencies of the new lines, Kayser concluded that they have this characteristic, and constitute a new hydrogen series. If these lines can be produced in laboratory experiments, im- portant information as to stellar temperatures and pressures is likely to be obtained. At low temperatures wafer vapor gives an absorption- spectrum rich in lines which are chiefly confined to the red region; these constitute a large number of the terrestrial Fraunhofer lines, and are referred to under nitrogen ; they are strongest when the sun is low in the horizon, as its rays have then to traverse a considerable layer of the atmosphere. When the latter is saturated with moisture, a " rain-band " is visible with the help of a spectroscope of low dispersive 1 Bull. Acad. St. Petersb. (1880) 11, 307; (1884! 12, 203. Mem. Acad. St. Petersb. (1882) 30, No. 7; (1883) 31, No. 14. W. A. 15, 45. See also* H. C. Vogel, P. A. (1872) 146, 569. Wtillner, P. A. 135, 497; 137, 337; 144, 481. W. A. 14, 355. Seabroke, P. M. [4] 43, 155. Balmer, W. A_ (1885) 25, 80. 2 Astrophys. J. (1897) 5, 92. Science (1897) 5, 726. 8 Astrophys. J. (1897) 5, 95. Science (1897) 5, 726. 1 30 SPEL'TR UM ANA L YSIS. power: it consists of bands composed of water- vapor lines, and situated between the red end and the ZMine. The presence of the rain-band has been used by Piazzi-Smyth, 1 Capron, 2 Grace and others as a means of prognosticating rain. Janssen 8 investigated the absorption-spectrum of steam contained in long tubes under considerable pressure, and Schonn 4 states that pure water exhibits an absorption-band. An emission-spectrum consisting of numerous lines in the ultra-violet is obtained by burning hydrogen in air, or passing it through the electric arc; it has been measured by Huggins, and also by Liveing and Dewar, 5 who distinguished five series of lines. The first, between wave-length 3268.2 and 3063.7, contains about 1 16 lines; the second, from 3057 to 2812, comprises 180 lines; the third, from 2807 to 2609, contains 141 lines; the fourth series, from 2606 to 2450, has 88 lines; and the fifth series includes 79 lines between wave-length 2449 and 2268. Elementary line-spectrum: [C or Ha] 6563.04 [F or H/J] 4861.49 [G or H^] 4340.66 [h or US] 4101.85 [H] 3970.25 []38S9.i5 [<5] 3770. 7 [e] 3752.05 [o] 37II-9 (Ames). m 3835.5 [C] 3 734- 1 5 M 3798.0 [77] 372 1. S Compound line-spectrum: 6135-5 5938.9 5013-15 6121.9 5931-8 4973-3 6081. o 6070.7 5888.9 5884.5 4934.5 4928.8 6032.1 5813.0 4861.49 6018.5 5084.9 4719.2 5975.8 5055. *2 4683.95 1 N. 26, 551. 2 The Observatory, 1882, pp. 42, 71. 3 C. r. 63, 289. 4 P. A. Suppl. Bd. (1878) 9, 670. W. A. 6, 267. See also for the spec, trum of water Huggins, P. R. S. 30, 576. C. r. (1872) 74, 1050. Liveing and Dewar, P. R. S. 30, 580; 33, 274. P. T. (1888) 179, 2. Lecoq de Boisbaudran, C. r. 74, 1050. Deslandres, A. c. p. (1888) [6] 14, 257. C. r. 100, 854. Absorption spectrum, Janssen, C. r. 54, 1280; 56, 538; 60, 213. Russell and Lapraik, N. (1880) 22, 368. Soret and Sarasin, Arch. Sc. phys. et nat. (1884) 11, 327. C. r. 98, 624. Ewan, P. R. S. (1895) 57, 126. 6 P. R. S. 35, 74. SPECTRA OF THE ELEMENTS. 4634.15 4461.1 4177.25 4I7L35 3990.15 3970.25 3796.8 3684.3 4412.35 434066 4212.65 4101.85 4079.0 4069.75 3889.15 3871-8 3863.3 3674.5 (Hasselberg, Ames). INDIUM. 4205.2 4195-9 4067.0 4062.6 3835.6 3804.9 This element was discovered in 1864 by Reich and Richter ' by means of its flame-spectrum, which consists of an indigo-blue and a violet line. The visible portion of the spark-spectrum has been measured by Clayden and Heycock, 2 and a small part of it also by Thalen, 3 whilst Hartley and Adeney 4 investigated the ultra-violet region. The arc-spec- trum has been examined by Liveing and Dewar; 6 it contains only the two lines above mentioned, which were accurately measured by Kayser and Runge. 8 Indium furnishes a remarkable example of the simplicity of the arc-spectrum as compared with that of the spark; the latter contains numerous lines both in the visible and ultra-violet regions which are absent from the former. Flame-spectrum 7 : 4511.44 4101.87 Arc-spectrum : 4511.44 4101.87 2932.71 2753-97 2560.25 2521.45 3258.66 2714.05 2460. 14 3256. 17 303946. 2710.38 2601.84 2389.64 2340.30 Spark-spectrum: 6907.6 6193.9 6096.0 5821.0 5645.0 5251.0 4680.9 4656.9 4638.8 4532.8 4511.44 4253.7 4101.87 4072.3 4064.2 4033-4 3853.5 3835.3 3258.66 3256.17 3039.46 2932.71 2890.2 2710.38 2560.25 2527-5 2521.45 2460.14 2389.64 235T.7 2306.8 J. pr. Chem. 89, 441. P. M. (1876) 5, 387. N. A. S. U. (1868) [3] 6. P. T. (1883) 176, 63. P. R. S. (1879) 28, 367. A. B. A. 1892. Lecoq de Boisbaudran, Spectres lumineux (Paris, 1874). Mtlller, P. A. 124, 637. 132 SPECTRUM ANALYSIS. IODINE. The more important investigations of the line-spectrum of iodine are those of Pliicker and Hittorf, 1 and of Salet; 2 the latter passed sparks from Leyden jars, or a Holtz machine, through a Geissler tube containing iodine: the tube was not closed until the iodine vaporized. There are no new deter- minations. The spectrum obtained with a condenser in the circuit differs from that produced by the continuous dis- charge. 3 The violet iodine vapors produce an absorption- spectrum extending over the red and green, but not over the blue and violet regions; it consists of numerous slender lines grouped into bands, and has been accurately measured by Hasselberg. 4 H. W. Vogel 6 has examined the absorption- spectra of various solutions of iodine, whilst Gernez * has studied those of iodine chloride and iodine bromide. The rule that the coefficient of extinction of solutions of a colored substance changes with the concentration, is found by E. Thiele 7 not to apply to iodine; he investigated the absorption-spectrum of solutions of varying strength in the same solvent. . Previous exceptions to this rule had all been electrolytes, and their abnormal behavior had been thought to be due to electrolytic dissociation, but iodine is a non- electrolyte. 1 P. T. 155, 24. 2 Spectroscopie (Paris, 1888). C. r. 74, 1249; 75, 76. A. c. p. [4] 28, 29. See also Pliicker, P. A. (1859) 107, 638. Ciamician, Wien. Ber. [2] 78, 877. Wullner, P. A. 120, 158. Mitscherlich, P. A. 121, 474. 3 Trowbridge and Richards, .Amer. Jour. Sci. (1897) [4] 3, 117. P. M. 43, 135. 4 Mem. de 1'Acad. St. Petersb. (1888) [7] 36. See also Daniell and Miller, P. A. 28, 386. Morghen, Beiblatter, 8, 822. Thalen, Le spectre d'Absorption de la vapeur d'lode (Upsal, 1869). 6 Ber. 11, 919. Monatsber. Berl. Akad. 1878, p. 417. C. r. 74, 466. 7 Zeitschr. phys. Chem. 16, 147. Ber. 28, R 720. SPECTRA OF THE ELEMENTS. 133 Spark-spectrum : 6258 6211 6126 6079 5953 (5791 5774 5761 5739 5712) (5689 5674) 5625 (5495 5462 5534 5404) (5345 5337) 5244 5163 5016 4866 (4678 4669) 4634 Absorption-spectrum of iodine vapor: Group Group Group Group Group Group 6316 6272 : 62726234 : 6234 6191 : 6191 6149 : 61496111 : 6m 6069 : 6316.51 6291.94 6253.07 6233.93 6190.97 6111.25 6108.87 6298.29 6291.46 6252.96 6229.68 6153.08 6069.31 6297. 6289. 6237. 6212. 6149. 76 83 72 4i 48 6294. 6272. 6210. 75 42 18 6294.25 6191.87 Group 6069 6031 : 6063. 49 6047. 82 6046. 87 6045. 94 6042.81 6035. 82 6034. 83 b (6033. 40 6033. 05) b (6031 .92 6031. 58) Group 60315992 : 6030.99 6024. 38 6020. 38 6018. 37 6008.85 (5994- 65 5994- ^2) 5993- 89 5993- 03 Group 59925976 : Group 59555917 : (5948.83 5948. 62) b(5922. 53 5922. 04) b(5 9 2i 77 5921. 24) 5(5921. 00 5920. 58) 5920.00 5919 75 59I9- 36 59i9- II 5917. 55 Group 59175881 : 5885. 00 5884. 74 5884. 10 5881. 17 Group 58815846 : 5866. 9i 5859- 85 5856. 49 (5850. 5i 5850 .22) 5848.57 5847-08 5846. 54 5846. 22 Group 58465811 : 5815- 40 5812. 66 5811. 65 Group 58115778: (5805. 86 5805. 63) (5798. 69 5798. 45) (5798 .14 5797- 93) 5793- 47 5778. 87 5778. 62 5778 .28 Group 57785746 : 5746. 21 Group 57465715 : 5745- 92 5720. 60 5715. 45 Group 57155684: 57I4- 92 5712. 24 5708. 38 5698. 70 5697 .84 5693- 05 5685. 09 5684.54 Group 56845655 : 5683. 08 5678. 59 5676. 82 5658. 98 Group 56555626 : 5654. 71 5649- 61 5647. 68 5646. 72 b(5646 .14 5645- 50) 5640. 90 5640.00 5639- 15 b(5638 .6 4 5638. 23) 5637- 36 (5628. 90 5628. 35) 5627.97 5627. 19 5626. 50 Group 56265599 : (5620. 59 5620. 33) 5616. 50 5614. 53 5602 .98 5601. 8r 5699' 14 Group 55995587: 5588. 98 5587.56 Group 55875560 : 5585. 10 (5577. 63 5577- 42) 5574- ii 5572 71 556i. 58 556o. 44 556o. 25 5559- 57 Group 55605533 : 5557- 17 5553- 61 5549- 40 b(5547. 92 5546 .96 5546.07) 5545. 57 5542. 37 5540.91 5533 37 134 Group SPECTRUM ANALYSIS. 5533 5507:b(553i-75 553i-io) 5527-58 5526.38 b(5522. 25 5521.79) 5521. 34 5520. 33 Group 55075482 : 5505. 19 5504. 95 55oo. 43 5498. 32 5497. 81 (5497- 51 5497-15) 5496. 36 5490. 37 5489- 67 5488. 95 0(5483.00 5482. n) Group 54825457 : 5473-55 b(5473- 12 5472. 67) 5468. 38 5466. 76 (5457- 90 5457-08) Group 54575434 : 5438- 43, 5434-03 Group 54345411 : 5430.71 5425. 7 6 54I9- 78 54io- 85 5416. 99 54I4- 28 5412 31 541'- 66 Group 54115389: b 5410.75 5404- 9 6 5404.04 b(5393- 9 1 5393- 44) 5390. 85 b (5390. 21 5389. 01) Group 53895367 : 538i. 90 538o. 93 5378. 05 5377- 32 5375- 20 5374- 38 5369- 74 5369- 20 5368. 5i 5368. 01 Group 53675347 : 5364. 76 5358. 81 5357- 91 5356. 63 5355- 89 5353- 28 5350. 56 5349- 87 5348.06 5347- 35 Group 53475327 : 5346. 24 5343- 12 5341- 43 5340. 14 5333- 73 5333- 10 5330.97 5329. 32 Group 53275308 : Group 53085291 : Group 52915273 : 5290.72 Group 52735255 : 5272. 75 Group 52555240 : Group 52405224 : 5224. 10 Group 52245209 : 5215- 83 5209.46 Group 5209-5195 : 5195. 22 Group 51955182: Group 5182 5168 : 5181. 9 6 5168. 65 Group 51685156: 5161. 45 5156. 16 Group 51565145 : 5144- 71 IRIDIUM. There are no accurate measurements of the spectrum of this element. Kirchhoff 1 observed three faint lines of wl t 6348.1, 545 O> 6 and 5300 6, and Lockyer 2 has measured six in the arc-spectrum between wave-length 4000 and 3900. H. W. Vogel 3 has described the absorption-spectrum of ammonium iridio-chloride. Arc-spectrum : 3992 2 3976.0 3945- 3934-7 3915.2 3902.5 1 A. B. A. 1861. 2 P. T. 1881, Pt. 3. 8 Prakt. Spectralanal. (Berlin, 1889). SPECTRA OF THE ELEMENTS. 13$ IRON. The spectrum of iron is the richest in lines; they are dis- tributed over every part of the field, and therefore, like the Fraunhofer lines, are excellently adapted for purposes of orientation in spectroscopic observations. Omitting the older and less accurate measurements, the following investi- gations are of greatest importance : Thalen ' made a careful comparison of the lines in the iron-carbon arc with those of the solar spectrum shown in the atlases of Angstrom, and of Fievez and Vogel, and investigated those between 760^ and 4OO,w/u. In the ultra-violet Cornu ' photographed the more prominent lines between 4iOju/w and 295^, and Liveing and Devvar 3 caiefully extended this work to the region between 295;^ and 230^^; they also, as did Hartley and Adeney, 4 investigated the spark-spectrum of iron. All these measurements were based on Angstrom's erroneous determi- nation of the wave-length of the ZMine, so that a reinvestiga- tion of the subject was very desirable. Kayser and Runge * undertook this task, and usually attained an accuracy of 0.02 Angstroms; they measured more than 4500 lines, and on comparing them with Rowland's solar atlas between 520^^ and 320;.^, they were unable with certainty to detect a single line which does not appear in the solar-spectrum. The absrption-spectra of solutions of various compounds of iron are not specially characteristic; they have been studied by N. A. S. U. [3] 6. Le spectre du fer. 1884. Spectre normal du soleil (Paris, 1881). P. T. 174, 210. P. R. S. 32, 402. P. T. 1884. A. B. A. 1888, 1890. See also Huggins, P. T. (1864) 154, 139. 'Kirch- hoff, A. B. A. 1861. Angstrom, Recherches sur le spectre solaire, 1868. Mascart, Ann. de 1'ecole normale (1866), 4. Secchi. C. r. (1873) 77, 173. Lecoq de Boisbaudran, Spectres lumineux (Paris, 1874). Lockyer, P. T.. 164, 479. H. W. Vogel, Prakt. Spectralanal. (Berlin, 1889). UNIVERSITY 136 SPECTRUM ANALYSIS. H. W. Vogel, 1 and are referred to in the following chapter. Roscoe," Lielegg, 3 Marshall Watts, 4 and others have investi- gated the spectrum of the flame observed during the manu- facture of steel by the Bessemer process, and more recently it has been thoroughly examined by Hartley. 5 Watts believed that the green bands are produced by manganese oxide, and that their disappearance marks the instant when the iron is completely decarbonized, and the blast of air should be stopped. Before the application of the spectroscope for this purpose it was exceedingly difficult to determine the precise moment, since- experience was the sole guide and often proved untrustworthy. Hartley's exact investigations have shown that the phenomena of the Bessemer flame is much more complex than was previously divined. This is owing to the super- position of bands of manganese, carbon, carbonic oxide, and possibly also of those of manganese oxide, and of the lines of iron, manganese, potassium, sodium, lithium, and hydrogen. The bands of manganese are to some extent obscured by the strong continuous spectrum of the carbonic oxide flame, and by the bands of carbon. The cause of the nonappearance of the lines in the spec- trum at the beginning of the " blow " is the comparatively low temperature at this period, and the free oxygen, which escapes with carbon dioxide, giving a gaseous mixture con- 1 Ber. (1875) 8, 1537. See also Miiller, P. A. (1847) 72, 67. Ewan, P. K. S. (1895) 57, 140. 2 P. M. P. S. 1863. 3 Sitzungsber. Wien. Akad. 1867. 4 P. M. [4] 34, 437; 38, 249- 5 P. T. (1894) 185, 1041. P. R. S. (1895) 59, 98. Journ. Iron and Steel Inst. (1895) No. TI. See also Kohn, Dingier polyt. J. (1864) 175, 296. Silli- man, P. M. 41, i. Tunner, Dingier polyt. J. (1865) 178, 465. Brunner, Oester. Zeitschr. fvirBerg-und Hiittenwesen (1868), 16, 226. Kuppelwieser, Ibid. (1868) 16, 59. v. Lichtenfels, Dingl. polyt. J. 191, 213. Spear Parker, C. N. 23, 25. Wedding, Zeitschr. f. Berg-. Hiitten- und Salinen- wesen (1869), 27, 117. Greiner, Revue universelle (1874), 35, 623. SPECTRA OF THE ELEMENTS. 137 taining too small a proportion of carbonic oxide to produce luminosity. During the " boil " the luminosity of the flame is due to the combustion of highly heated carbonic oxide, and also to the presence of the vapors of iron and manganese. The disappearance of the manganese-spectrum at the end of the " fining " stage is primarily caused by the carbonic oxide, which escapes from the converter, being reduced in quantity owing to the diminished supply of carbon in the metal. When the last traces of carbon are gone, so that air can escape through the metal, the blast instantly oxidizes any manganese, either in the metal or in the atmosphere of the converter, and also some of the iron. The temperature must then fall with great rapidity. Arc-spectrum : 6678.23 6663.38 6633.92 6609.35 6594-11 6593.14 6575.17 6569.46 6546.47 6518.62 6495.20 6431.06 6421.55 6420.17 6411.86 6408.23 6400.27 6393-81 6380.95 6337.07 6335.55 6322.91 6318.22 6302.73 6301.71 6297.99 6291.18 6270.44 6265.34 6256.57 6254.45 6252.76 6246.53 6232.90 6230.94 6219.49 6215.36 6213.63 6200.53 6191.77 6180.42 6170.69 6157.96 6141.93 6137.91 6136 85 6128.11 6103.45 6102.40 6078.71 6065.71 6056.21 6042.31 6027.27 6024.28 6020.35 6008.78 6003.28 5987.29 5985-04 5983.9? 5977.12 5975-58 5956.92 5953-ot 5934.83 5930.29 5916.47 5914.38 -905.94 5862.58 5859.91 5816.59 5782.35 5775.30 5763.22 5753-33 5731.98 5718.10 5709.61 5701-77 5686.66 5662.75 5659 06 5638.52 5624.77 5615.88 5603.17 558699 5576.32 5573.07 5569.85 5565-83 5555-03 = 506.98 5501.69 5497-70 5476.89 5474-13 5463-49 5455.83 5447.16 5445.28 5434-74 5429.81 5424.23 5415.42 5411.20 5405.99 5404.42 5400.67 5397.32 5393.38 5383 58 5371.67 5370.17 5367-67 5365.08 5353-59 5341.21 s-uo. 16 5333-09 5328.71 5328.21 5324.37 5307-54 5302.52 5283.80 5281.97 5273.55 5270.52 [E 2 ]5269.72 5266.73 5263.42 5250.81 5242.66 5233.12 5230.01 5227.39 5227.08 5216.43 5208.80 5202.48 5195.09 5192.53 5191.68 5I7L78 5169.07 5167.57 5162.45 5153-34 5151-02 ?T48.42 5139.65 5139.44 5137.56 5133.70 5125.30 5123.88 5110-57 5107.82 5105.73 J38 SPECTRUM ANALYSIS. 5098.83 5083. 50 5079.91 5079.42 5068.95 5065.15 5051.78 5050.01 5041.91 5015-13 5012.21 5006.30 5005.90 5002.08 4982.73 4966. 29 4957-50 4957.48 4938.99 4920.68 4919.19 4891.68 4890.94 4878.41 4872.31 4871.49 4859.94 4789.80 4737.00 4707.51 4691.59 4679.03 4668.36 4667.62 4654.76 4647.60 4638.19 4637.68 4625.25 4619.46 4611.47 4607.85 4603.09 4598.32 4592.82 4556.28 4548.01 453L3I 4538.84 4525.32 4494- 74 4484.42 4482.40 4476.25 4469.58 4466.75 4461.83 4459-29 4454-55 4447.90 4443-35 4442.52 4433.37 4430.79 4427.49 4422.74 4415 30 4408.59 4407.85 4404.94 4401.51 4391.14 4388.62 [d]4383.72 4376.10 4369.94 4367.73 4352.90 4337.20 [f]4325 94 4315.23 4309.55 [G 14308. 07 4305-63 4299.44 4294.31 4285.62 4282.54 4271 92 4271 35 4268.02 4260.67 4250.96 4250.30 4247.65 4245.40 4239'95 4239.03 4236.14 4233.81 4227.65 4225.66 4224.32 4222.39 4219.52 4217.74 4216.33 4210.53 4204. 12 4202.18 4199.26 4198.47 4196.36 4I95-5I 4191.62 4187.97 4187.22 4185.06 4182.51 4181.91 4177.71 4176.67 4175.76 4175.03 4172.86 4172.25 4171.04 4158.94 4I57.95 4156.93 4i55.oo 4154-62 4154.09 4149.49 4147.79 4144.01 4143.58 4137-11 4134.82 4133-01 4132.20 4127.73 4122.64 4121.97 4118.72 4114.60 4109.93 4107-65 4104.25 4100.88 4098-31 4096. 1 1 4085.43 4085.12 4084.64 4079.96 4078.46 4076.77 4074.92 4071.90 4070.90 4068.12 4067.41 4067.09 4063.75 4062.60 4057.96 4045.97 4034.64 4033.21 4030.89 4022.02 4014.68 4009.85 4005.31 3998.22 3997-54 3986.32 3984.08 3981.91 3977-90 3971.48 3969.39 3956.82 3952.76 3951.30 3950.10 3948.92 3942.58 3941.03 3935-97 3933.80 3930-37 3928.06 3923.05 3920.41 39I7.34 3916.88 3906.63 3904-05 3903-11 3899.85 3898.10 3895.80 3888.68 3887.22 3878.71 3886.42 3878.17 3873.93 3872.66 3867.38 3865.70 3860.04 3859-39 3856.52 3852.76 3851.01 3850.16 3847.01 3843-41 3841.24 3840.58 3839-43 3836.53 3834.42 3827.97 3826.02 3824.63 [L] 3 820.56 3815.98 3813.17 3806.89 3805.49 3799.70 3798.66 3797.70 3795-15 3790.27 3788.03 3779-63 3767.33 . 3765.71 3763.94 3758.38 3749-62 3748.41 3746.05 3745-71 3743-51 373?.48 3737.26 3735.49 ,.3735.oi 3732.55 [M]( 3 72 7 .77 3727.17) 3724-b5 3722.69 3720.05 3716.59 3709.37 3708.07 3704-63 3701.24 3694.17 3687.81 3687.62 3686.14 3082.39 SPECTRA OF THE ELEMENTS. 139 3669.69 3659-64 3651.65 3647.95 3640.57 3631-59 3623 37 3622.19 3621.65 3618.89 3617-98 3610.33 3609.03 3606.87 3605.66 3594-75 3587.14 3586.24 3584.82 [N]358i.34 3574.04 3572.16 3570.23 3565.53 3553.69 3557-03 3555-08 3545.78 3542.24 3241.20 3536.69 3533-34 3527-94 3526.55 3526.29 3521.40 3513.97 3497-99 3497-27 3490.72 3489.78 3476.85 3475 56 3471-44 3465.97 3460.06 3452-39 3452.03 3450.44 3447-41 3445.26 3444.02 [OJ3441.13 3440.70 3428.30 3427-24 3426.75 3426.48 3425-12 3424.40 3422.73 3418.61 3417.96 3415.65 3413.26 3410.30 3407.57 3406.96 3404.44 3402.37 3401.64 3399.43 3394-69 3392.76 3384-07 3380.21 3379-15 3378.78 337091 3369-66 3366.92 3355-33 3348.03 3342.39 3340.68 3337-77 3329-04 332388 3314.87 3310.57 3307.37 3306.48 3306.10 3298.24 3292-74 3292.17 3291.14 [QJ3286.87 3280.41 3274.09 3271-11 3265.72 3257.73 3254-5I 3251.32 3248.35 3244.31 3239.54 3234-II 3231.07 3227.92 3225.91 3222.16 3219-91 3219-70 32I7.53 3216.07 3214.14 3212.12 3205.49 3200.58 3199.66 3197-08 3T93.4I 3192-89 3191.81 [RJ318034 3178.12 3175.54 3171-48 3166.59 3166.01 3162.08 3160.74 3158.03 3I57.I9 3153.35 3151-42 3I44.IO 3142.58 3I34.25 3132.65 3126.29 3125.76 3119.62 3116.73 3102.80 3100.78 [S 2 ]3ioo.42 3100.06 3098.29 3093.96 3091.71 3083 84 3075.82 3067.35 3059.23 3057.54 3055.39 3053.I7 [s]3047.70 3045.20 3042.79 3042.17 3041-87 3040.58 3037.49 3037.41 3031.78 3031.35 3030.28 3026.61 3025.96 3024.15 [T]( 3 02i.i9 3020.76) 3019.09 3017-75 3016.29 3011.61 3009.70 3008.23 3007.34 3003.18 3001.04 3000. 60 2999.63 [t]2991.52 2991.82 2990.48 2987.40 2985.69 2984.96 2983.67 2981.99 2981.57 2980.66 2976.23 2973.36 2973-25 2970.22 2969.56 2967.00 2965.38 2960.11 2957.61 2957.48 2957.42 2954.06 2953-90 2953.61 2950.38 2949.32 2948.56 [U]2947. 9 9 2947.. i 2944-53 2941.46 2937-94 2937.02 2929.13 2926.69 2925.47 2923-96 2923.43 2920.79 2918.14 2914.37 2912.27 2909.60 2909.00 2907.60 2902.05 2901.48 2899.52 2898.55 2895.14 2894.60 2892.59 2887.91 2886,41 2883.82 2881.68 2880.87 2877.40 2874.27 2872.41 2869.41 2866.71 2863.95 2863.48 2862.59 2858.99 2853.84 2852.22 2851.89 2850.72 2848.80 2846.90 2845.66 2844.08 2843.74 2840.53 2840.09 2838.23 2835-54 2832.53 2828.90 2825.78 2825.67 140 SPECTRUM ANALYSIS. 2824.45 2823.37 2819.38 2817.58 2815.61 2813.39 2808.40 2807.08 2804.59 2803.71 2801.18 2798 36 2797.85 2795-61 2795.03 2794.80 2792.47 2791.87 2791.54 2789.90 2788.20 2788.08 2783-78 2781.95 2779.37 2778.92 2778-32 2778.18 2774.79 2773.3I 2772.21 2769.40 2767.62 2767.02 2764.44 2763.20 2762.85 2762.11 2761.88 2760.99 2759.89 2757.94 2757.41 2756.43 2755.83 2754.51 2754.12 2753.77 2753.40 2750.98 2750.24 2749.64 2749-45 2749.26 2747.67 2747.08 2746.57 2745-16 2744.63 2744.16 2743.66 2743.26 2742.48 2739.62 2737.40 2737.05 2735.74 2735.64 2735.54 2734.42 2733.67 2730.82 2728.93 2728.14 2727.64 2726.23 2725 oo 2723 67 2720.99 2720.31 2719.54 2719.12 2718.54 2714.51 2711.74 2710.64 2708.67 2706.68 2706.10 2704.09 2699.21 2697.11 2696.44 2696.15 2690.15 2689.95 2689.31 2680.56 2679.16 2673.31 2669.60 2668.00 2667.08 2666.97 2666.46 2664.77 2662.16 2661.32 2660.51 2656.88 2656.24 2651.81 2647.67 2645-55 2644.08 2641.77 2635.90 2631.40 2631.12 2628.39 2625.75 2623.61 2621.75 2620.50 2617.70 2615.53 2613.94 2611.96 2607.19 2605.80 2604.93 2599.49 2598.46 2594.23 2593-78 2591.64 2588.14 2585.95 2584-63 2582.53 2579-95 2578.04 2576.77 2576.23 2575.86 2574.46 2572.85 2570.59 2569.76 2566.99 2563.56 2562.61 2560.68 2556.95 2556.41 2553-35 2551.22 2549.68 2548.79 2547.09 2546.29 2544-86 2544.05 2542.23 2541.06 2539.01 2537.24 2536.93 2535.70 2533.89 2532.40 2530.82 2529.43 2528.60 2527.50 2527.33 2526.33 2525-51 2525.14 2524-35 2523-79 2523.22 2522.92 2522.00 2521.12 2518.19 2517.79 2517.28 2516.22 2514-41 2512.41 2511.08 2510.91 2508.81 2508.02 2507.01 2505.67 250303 2502.56 2501.90 2501.20 2498.99 2497.91 2497-18 2496.63 2496.04 2493.30 2491.24 2491.01 2489.07 2488.24 2484.28 2483.34 2480.28 2480.04 2472.85 2476.80 2474.91 2473-18 2472.86 2472.43 2469.00 2467.83 2466.84 2465.23 2462.74 2461.31 2460.40 2458.81 2457.68 2453.56 2447.84 2444.61 2443-97 2442.65 2440. 28 2439.85 2439.39 2438.30 2436.48 2435.07 2431-11 2430.19 2429.56 2424.23 2421.82 2413.39 2411.19 2410.60 2406.74 2404.96 2404. 5 1 2399.31 2395.72 2391.56 2388.73 2384.51 2383.27 2382.12 2380.85 2379.41 2375-33 2373.82 2370.59 2368.69 2366.69 2364.91 2362.14 2360.40 2360.09 SPECTRA OF THE ELEMENTS. 2359.20 2354.96 2348.39 2344.40 2344. 1 2 2343.56 2338.11 2332.86 2331.41 2327.45 2320.43 2313.20 2309.08 2303-55 2298.25 2297.88 2293.98 2291.21 2289.06 2230.13 2227.81 2214.70 LANTHANUM. The spark-spectrum of lanthanum is rich in lines which have been measured by Kirchhoff, 1 and more especially by Thalen." Spark-spectrum: 6393.5 5770-0 5376.5 5123.0 4330.5 4217.0 4042.5 6250.0 5974-0 5930.0 5805.5 5795-0 579L5 5788.0 5674.0 5632.0 5588.0 5501.5 5455-5 538i.o 5381.5 5340.5 5303-5 5302.8 5302.0 5188.5 5183.5 517^0 4526.5 4525-0 4523.0 4431.0 4428.0 4385-0 4383 4322.5 4295.5 4286.5 4268.5 4263.5 4238.5 4235.5 4196.5 4192.0 4152.0 4142.5 4122.0 4086.5 4077.0 403L5 3947-0 LEAD. The arc-spectrum is obtained by the use of leaden elec- trodes, preferably in an atmosphere of hydrogen, so as to avoid the production of the oxide bands which were observed by Plucker and Hittorf 3 on passing sparks through a Geissler tube containing lead chloride vapor. The arc-spectrum differs materially from that of the spark; it has been investigated by Liveing and Dewar, 4 and more recently by Kayser and Runge. 5 Lockyer and Roberts 8 observed that, at low tem- peratures, lead vapor produces an absorption-spectrum in the red and blue. All lead compounds exhibit the channelled oxide-spectrum consisting of the bands of \ 5905, 5685, 5611, and 5461; they shade off towards the red, and have been described by Mitscherlich, 7 Plucker and Hittorf, and A. B. A. 1861. Konffl. Svenska Vetensk. Akad. Handl. (1874) 12, No. 4. See also Bu sen. P. A. 155, 366. Cleve, C. r. 95, 33. P. T. 155, 25. P. R. S. (1879) 29, 402. P. T. (1882) 174, 18-7. A. B. A. 1893. P. R. S. 23. 344. Lockyer, P. T. 163, 253, 369. P. A. 121, 468. 142 SPECTRUM ANALYSIS. Lecoq de Boisbaudran. 1 In the Bunsen flame they are too fugitive to be observed with certainty, but H. W. Vogel 2 has constructed an apparatus which permits of the addition of lead chloride vapor to a hydrogen or coal-gas flame, and renders them much more permanent. Arc and spark spectra: 6657.4* 6453.3* 6041.2* 6002.08 5875-1* 5608.0* 5547-2* 5373.4* 5201.65 5045-9* 5005.62 4387.3* 4246-6* 4057.97 3740.10 368360 3639.71 3572.88 3262.47 3240.31 3220.68 2873.40 2833.17 2823.28 2802.09 2697.72 2663.26 2650.77 2614.26 2577.35 2476.48 2446.28 2443.92 2428.71 2411.80 2402.04 2393.89 2332.54 2247.00 2237.52 2175.88 2170.07 2115.1 2112.0 2088.5 LITHIUM. Lithium salts are dissociated in the Bunsen flame and exhibit two lines belonging to the metal : the one, A = 6708.2, is very bright and deep red; the other, A = 6103.8, is fainter and orange-colored. Other lines are visible in the spark a'nd arc spectra. Kayser and Runge 3 measured eighteen, and Liveing and Dewar 4 observed two additional ones in the ultra-violet. Arc and spark spectra: 6708. 2f 6103. 77f 4972.11 4602.37 4132.44 3915.2 3232.77 2741.39 MAGNESIUM. The bibliography shows how frequently the magnesium- spectrum has been subjected to exhaustive investigation; it 1 Spectres lurnineux (Paris, 1874). 2 Prakt. Spectralanal. (Berlin, 1889). See also Thalen, N. A. S. U. (1868) [3] 6. Kirchhoff, A. B. A. 1861. Huggins, P. T. 154, 139. Hartley and Adeney, P. T. 175, 163. 3 A. B. A. 1890. 4 P. R. S. 28, 367, 471 ; (1880) 30, 93. P. T. (1883) 174, 215. See also Kirchhoff and Bunsen, P. A. 110, 167. Kirchhoff, A. B. A. 1861. Huggins, P. T. 1864, p. 139. Miiller, P. A. 118, 641. Ketteler, P. A. 104, 390. Wolf and Diacon, C. r. 55, 334. Riihlmann, P. A. 132, i. Thalen, N. A. S. U. (1868) [3] 6. Lecoq de Boisbaudran, Spectres lurnineux (Paris, 1874). * Visible only in the spark-spectrum. (Thalen.) f Visible also in the flame-spectrum of lithium salts. SPECTRA OF 7^HE ELEMENTS. 143 is obtained by burning the metal in air, by passing sparks through a solution of a magnesium salt, or between electrodes of the metal, and by the combustion of the metal ini the carbon arc. Liveing and Dewar 1 examined the flame, spark, and carbon arc spectra in atmospheres of \arious gases; the spectra are identical, although the lines may differ in bright- ness. They also observed a number of oxide bands of A = 5006-4934 and 3865-3720; these shade off in the violet. A band between A= 3634 and 3621 is also visible in the oxyhydrogen flame. They sta'e that when magnesium is burnt in hydrogen the elements combine, and the resulting product exhibits a band-spectrum between /I = 5618 and 4803. Kayser and Runge 2 have made more recent and accurate measurements of the arc-spectrum, using magnesium powder, or wire, and carbon poles; they were unable to detect the bands of the oxide or hydrogen compound, but the former were plainly visible on burning the metal in air. E. Becquerel 3 found the infra-red lines A = 8990, 10,470, 12,000, and 12,120 in the arc-spectrum. Magnesium compounds are not dis- sociated in the Bunsen flame. Spark and arc spectra: 5528.75 5183.84* 5172.87* 5i67-55* 47O3-33 4571-33* 4352.18 3838.44 3832.46* 3329.51* 3336.83 3332.28 3330.08 3097.06 3093.14 3091.18 2942.21 2938.67 2852.22* 2802.80 2795.63 2783.08 2781.53 2779.94 2778.36 2776.80 1 P. R. S. 28, 367 ; 30, 93 ; 32, 189. P. T. (1883) 174. 208. 9 A. B. A. 1891. 3 C. r. 96, 1218 ; 97, 72. See also Kirchhoff, A. B. A. 1861. Thalen, N. A. S. U. (1868) 6. Lecoq de Boisbaudran, Spectres lumineux (Paris, 1874). Cornu, Spectre normal du soleil (Paris, 1881). Bunsen, P. A. 155, 366. Fievez, 'Bull, de 1'Acad. R. de Belgique (1880), [2] 1, 91. Hartley and Adeney, P. T. (1884) 175, 95. * Also visible in the flame-spectrum. (Liveing and Dewar.) 144 SPECTRUM ANALYSIS. MAGNESIUM HYDRIDE. 1 Bands extending towards the red: 5619 5567 5514 5513 5512 5211 5181 4850 4804 MAGNESIUM OXIDE. 1 Bands extending towards the red : 5001 4991 4981 4970 MANGANESE. The spark-spectrum of manganese has been measured by Huggins,' 2 and by Thalen; 3 the arc-spectrum by Angstrom,* and Thalen, and in part by Lockyer, 5 and Cornu.' The spectra are not identical, and they have not been the subject of recent investigation except by Hartley, 7 who examined the oxyhydrogen flame-spectrum of manganese, and manganese oxide. Potassium permanganate exhibits the most characteristic absorption-spectrum of any manganese compound; it has been investigated by Gladstone, 8 Brewster, 9 and H. W. Vogel. 10 The last states that the spectra of the solid salt and of the solution are similar, but not identical; the spectrum of potassium manganate is quite different (comp. following chapter). The determination of potassium permanganate by spectro-colorimetric methods is described in G. and H. Krtiss* work on Colorimetry and Quantitative Spectrum Analysis. 1 Liveing and Dewar, P. R. S. 28, 367 ; 30, 93 ; 32, 189. 8 P. T. (1864) p. 139- N. A. S. U. (1868) [3] 6. Le spectre du fer, 1884. 4 Recherches sur le spectre solaire (Upsala, 1868). 6 P. T. (1873) 163, 270. 6 Spectre normal du soleil (Paris, 1881). 1 P. R. S. (1894) 56, 192. P. T. (1894) 185, 1029. See also Lecoq de Boisbaudran, Spectres lumineux (Paris, 1874). Liveing and Dewar, P. R. S. (1879) 29, 402. 8 Jour. Chem. Soc. 10, 79. P. M. [4] 24, 417. 9 7Wi] 5896.16* [Z> 2 ] 5890.19* 5688.43 5682.86 5I53-72 5I49-I9 49 8 3-53 4979-30 3303.07 3302.47 2852.91 2680.46 STRONTIUM. Mitscherlich obtained the line-spectrum by the use of the oxyhydrogen flame; it is also produced by the passage of sparks through a solution of the chloride, but the best effects are given by the volatilization of the chloride in the electric arc. This was the method employed by Kayser and Runge. 3 In the Bunsen flame the strontium haloid compounds chiefly exhibit their individual spectra, together with the band-spec- trum of the oxide, and the blue line, A = 4607.5, of the metal. 1 T. R. s. E. (1857) 21. 2 Kayser and Runge, A. B. A. 1890. Bunsen and Kirchhoff, P. A. 110 167. Kirchhoff, A. B. A. 1861. Attfield, P. T. 1862, 221. Rutherfurd, Silliman's Journ. [2] 35, 407. Huggins, P. T. 1864, p. 139- Wolf and Diacon, C. r. 55, 334. Miiller, P. A. 118, 641. Thalen, N. A. S. U.'(i868> [3] 6. Lecoq de Boisbaudran, Spectres lumineux (Paris, 1874). Lockyer P. R. S. (1879) 29, 140. Cornu, Spectre normal du soleil (Paris, 1881). Bunsen, P. A. 155, 366. Liveing and Dewar, P. R. S. 28, 367, 471 ; (1879) 29, 398, 402. E. Becquerel, C. r. 94, 1218 ; 97, 72. De Gramont, C. r. (1896) 122, 1411, 1443- 3 A. B. A. 1891. See also Bunsen and Kirchhoff, P. A. 110, 167. Kirchhoff, A. B. A. 1861. Mailer, P. A. 118, 641. Huggins, P. T. 1864, p. 139. Mascart, Annales de 1'Ecole normale (1866) 4. Thalen, N. A. S. U. (1868) [3] 6. Lockyer, P. T. 163, 639 ; 164, 311. E. Becquerel, C. r. 96, 1218 ; (1883^ 97, 72. Liveing and Dewar, P. T. 174, 217. Rydberg, W. A. (1894) 52, 119. * Visible also in the flame-spectrum. SPECTRA OF THE ELEMENTS. Arc and spark spectra: ^550. 53 6408. 65 6386.74 5543- 49 5540. 28 5535- 01 5522.02 5504.48 5486. 37 5481.15 5451. 08 5257. 12 5238. 76 5229.52 5225- 35 5222. 43 5156.37 4968. ii 4962. 45 4892.20 4876.35 4872.66 4868. 92 4855-27 4832. 23 4812. 01 4784. 43 4742.07 4722. 42 4678. 39 4607.52* 4531-54 4438. 22 4361. 87 4338.oo 435. 6o 4215. 66 4161.95 4077 .88 4030. 45 3705. 88 3547.92 4399-40 3475-01 3464-58 3380.89 3366. 43 3351. 35 3330.15 3322. 32 3307. 64 3301.81 2931. 98 STRONTIUM CHLORIDE. 1 Flame-spectrum: 6730 6599 6351 STRONTIUM OXIDE. 3 Flame-spectrum : c, ,t> 686'3.5t 6747-2f 6628f 6 499 t 6465* (6060 6/632) 4607.5 SULPHUR. Sulphur exhibits both a line- and a band-spectrum; the former was first observed by Seguin, 3 who passed sparks through a mixture of hydrogen and sulphur vapor; Plucker and Hittorf 4 volatilized sulphur in a Geissler tube, and em- ployed sparks from a Leyden jar. This general method is still in use; it suffers from the disadvantage that powerful sparks under a low pressure are apt to decompose any sulphur compounds in the glass, and so give rise to the sulphur- 1 Mitscherlich, P. A. 121, 459. Lecoq de Boisbaudran, Spectres lumi- neux (Paris, 1874). Bunsen, P. A. (1875) 155, 230. 2 Lecoq de Boisbaudran, Spectres lumineux (Paris, 1874). Eder and Valenta, Denkschriften Wiener Akad. (1893) 60, 473. 3 C. r. (1861) 53, 127. 4 P. T. 155, 13. * Blue line visible also in the flame-spectrum. f Bands sharply bordered towards the red, shading off towards the violet. % Middle of a band. 164 SPECTRUM ANALYSIS. spectrum when it is not desired. The band-spectrum is obtained by the passage of feeble sparks through a Geissler tube containing sulphur vapor. Salet 1 produced it by vola- tilizing sulphur, or one of its compounds, in a hydrogen flame, cooled by allowing it to impinge on a plate of metal or marble, on to the other side of which a stream of water was directed. This spectrum was mapped by Salet, and by Plucker and Hittorf, but the observations are limited to the visible region, and are too inaccurate to show more than the existence of the bands. The flame of burning sulphur exhibits a continuous spectrum which extends far into the violet. Line-spectrum: 5660.7 5640.3 5604.9 5562.4 5508.3 547^-4 5451.9 5439.0 5430.7 5342.6 5320.1 5215.4 5201.1 5143.3 5103.7 5033.3 5013.5 4994.7 4926.0 4919.4 4902.8 4885.4 4816.6 4715.8 4552.3 4525.5 4485.9 4464.7 Band-spectrum. a Bands sharply bordered towards the violet, but shading off towards the red (Salet): 5366 5221 5191 5089 5041 4991 4946 4841 4796 4656 4616 4471 TANTALUM. The lines of this element were too feeble for Thalen to measure, but Lockyer 3 observed eighteen of them in the arc- spectrum between A = 4000 and 3900. TELLURIUM. The line-spectrum of tellurium *is obtained by passing sparks between electrodes of the element, and has been meas- 1 A. c. p. [4] 28, 37. C. r. (1869) 68, 404. See also Hasselberg, Bull. Acad. imp. St. Petersb. (1880) 11, 307. Astronomy and Astrophysics (1893), 12, 347. Mulder, J. pr. Chem. (1864)91, 112. Ditte, C. r. 73, 559. Lock- yer, P. R. S. 22, 374. Gernez, C. r. (1872) 74, 803. Angstrom, P. A. 137, 300. C. r. 73, 368. Ciamician, Wien. Ber. 77, 839 ; 82, 425. Schuster, B. A. R. 1880, p. 272. Ames, Astronomy and Astrophysics (1893), 12. De Gramont, C. r. (1896) 122, 1326. 2 Salet, A. c. p. [4] 28, 37. C. r. (1874) 79, 1231. 3 P. T. 173, 561. SPECTRA OF THE ELEMENTS. I6 5 ured by Huggins, 1 and Thalen 8 in the visible region, and by Hartley and Adeney 3 in the ultra-violet. Salet 4 produced a band-spectrum by passing a discharge through a Geissler tube of hard glass containing tellurium; to facilitate heating, the tube was covered with metal. The spectrum consists of bands in the red, and channelled spaces in the green and blue; they are sharply bordered towards the violet, and shade off towards the red. The same spectrum is produced by volatilizing tellurium in a hydrogen flame. Gernez 5 investi- gated the absorption-spectra of tellurium chloride and bro- mide; they consist of channelled spaces, the former in the green and orange, the latter chiefly in the red and yellow. Spark-spectrum : 6438.2 '6047.2 6013.7 5974.1 5936.2 5782.0 5756.1 5707.6 5648. 1 5575-1 5489.1 5478.5 5448.5 5367.1 53II-0 5218.2 5I53-I 5104.9 4302.1 4275.0 4260.4 4221.7 4062.0 4054.8 4006.7 3984.5 3969-3 3948.7 3841.9 3736.1 3726.7 (3650.0 3645-1) 3617.8 3552-4 3520.9 3496.9 3456.7 344L9 3409-2 3384.1 3364-I 3354-6 3331-2 3309.3 (32820 3275.3) 3258.2 3248.7 3107.9 3073-1 3046.4 3017.0 2966.5 2941.2 (2894.7 2893.7) (2868.1 2860.4 2857.4) (2845-3 2840.4 2823.6) 2792-3 (2769.0 2766.9) 2710.6 (2697.0 2694.5) 2635.1 2544.1 2529.8 2505.6 2499.0 2473-6 (2448.2 2438.4) (2413-7 2411.8) (2404.1 2400.4) (23867 2384.2) 2370.7 2359.0 (2332.4 2325-9) 2321.4 2318.2 2295.4 (2281.0 2277.6) (2266.6 2261.0 2257.0) 2250.4 {2248.4 2247.7) 2243.7 2219.7 (2211.6 2209.9) (2192.6 2190.1) 2179.6 THALLIUM. Thallium salts, heated in the Bunsen flame, exhibit the characteristic green line of wave-length 5350.65. The visible portion of the spark-spectrum has been investigated by 1 P. T. 1864, p. 139. * N. A. S. U. (1868) [3] 6 (w.-l. 6438.2-5104.9). 3 P. T. (1883) 175, 63 (w.-l. 4302.1-2179.6). 4 A. c. p. [4] 28, 49. C. r. 73, 742. Spectroscopie (Paris, i 5 C. r. 74, 1190. See also Ditte, Ibid. (1872) 73, 622. 1 66 SPECTRUM ANALYSIS. Huggins, 1 and Thalen, 2 the ultra-violet region by Hartley and Adeney, 8 and Cornu. 4 The arc-spectrum has been measured by Liveing and Dewar, & and more recently by Kayser and Runge, 6 who usually obtained it from the metal, but occa- sionally used the chloride; they photographed it between the limits 63Oywju and 2io/f/w. The numerous lines in the spark- spectrum between 650;^ and 300;^ are almost all absent from the arc-spectrum. With the exception of the green line at 535/^, and a faint line at 553W, the rays of thallium, which are powerful, consist wholly of ultra-violet light. Arc and spark spectra: ' 5948.7* 5350.65f 5153-6* 5079-4* 5053.9* 4982.5* 4736.5* 4110.2* 3933.4* 3775.87 3529.58 3519.39 3229.88 2921.63 2918.43 2826.27 2767.97 2709.33 2665.67 2609.08 2580.23 2552.62 2379.66 2316.01 2237.91 THORIUM. The spark-spectrum of thorium has been measured by Thalen; 7 Rowland 8 records, in his table of wave-lengths, several additional bright lines which he obtained with the arc. Spark-spectrum: 5538 4382 .o .2 5446.9 4281.6 5375- 4278 5 .1 5350.67t 4273-1 4919. 3575. 9 4864. 3529. 5 55J 4393. 3519 2 34* 1 P. T. (1864) 154, 139. 2 N. A. S. U. (1868) [3] 6. 3 P. T. (1884) 175, 104. 4 C. r. (1885^ 100, 1181. 6 P. R. S. 27, 132. P. T. (1883) 174, 219. 6 A. B. A. 1892. See also W. A. Miller, P. T. (1862) 152, 861. Lecoq de Boisbaudran, Spectres lumineux (Paris, 1874). Crookes, P. T. 153, 277. P. M. [4] 26, 55. Bunsen, P. A. (1875) 155, 230, 366. Lockyer and Roberts, P. R. S. 23, 344. Ciamician, Wien. Ber. 76, 499. 7 N. A. S. U. (1868) [3] 6. 8 Astronomy and Astrophysics (1893) 12, 321. * Visible only in the spark-spectrum. (Thalen, Hartley and Adeney.) f Visible also in the flame-spectrum. \ In the arc-spectrum. (Rowland.) SPECTRA OF THE ELEMENTS. l6/ THULIUM. This element was discovered by Cleve in 1879, an ^ its spark-spectrum measured by Thalen. 1 The oxide is hardly perceptibly volatile in the Bunsen flame, but exhibits a dis- continuous spectrum consisting of the bands A = 6840 and 4760; the former corresponds with the absorption-bands shown by thulium salts, and Thalen describes an additional one at A = 4650. Kruss and Nilson consider that thulium, like other -metals of the rare earths, is not an element (comp. didymium). Spark-spectrum : 5962.5 5897.0 5676.0 5306.6 5034.3 4733-9 4615-8 4522.8 4481.8 4387-2 4360.1 4242.1 4204.6 4188.4 4107.3 4093.7 TIN. The spark-spectrum is obtained by the use of tin elec- trodes, or of a concentrated solution of a salt; it has been measured by Kirchhoff, 2 Huggins, 3 Thalen, 4 and Hartley and Adeney; 5 the arc-spectrum by Liveing and Dewar, 8 and Kayser and Runge. 7 The spectra differ considerably in the visible region, the latter only contains two lines, but four additional ones are included in the spark-spectrum, which also contains more lines in the ultra-violet. Salet " observed a band-spectrum of the oxide, and also a characteristic reddish yellow band at 6io/f/w, when the chloride is volatilized in a hydrogen flame; these results were confirmed by H. W. Vogel.' Ofversigt K. Vetensk. Forhandl. (1881) 40. A. B. A. (1861). P. T. (1864) 154, 139. N. A. S. U. (1868) [3] 6. Ibid. (1884) 175, 104. P. T. (1883) 174, 219. A. B. A. 1893. A. c. p. (1873) [4l 28, 68. 9 Prakt. Spectralanalyse (Berlin, 1889). See also Mascart, Ann. de I'Ecole normale (1866) 4. Lecoq de Boisbaudran, Spectres lumineux (Paris, $874). Lockyer and Roberts, P. R. S. (1875) 23, 344. 1 68 SPECTRUM ANALYSIS. Spark and arc spectra: 6453.3* 5799.0* 5631-91 5589.5* 5563.5* 4524.92 3801.16 3745- 7t 3595.9:}: 3352.3f 3330.71 3283.41 3262.44 3175.12 3034.21 3009.24 2913.67 2863.41 2850.72 2840.06 2788.09 2779.92 2706.61 2661.35 2658.3f 2643.6f 2631.9f 2594-49 2571.67 2558.12 2546.63 253L35 2495.80 2483.50 2429.58 2421.78 2408.27 2354.94 2334-89 2317.32 2286.79 2269.03 2267.30 2251.29 2246.15 2231.80 2209.78 2199.46 2194.63 2171.5 2151.2 2148.7 2141.1 2121.5 2113.9 2100.9 2096.4 2091.7 2080.2 2073.0 2068.7 2063.8 2058.3 2053.8 TITANIUM. The arc and spark spectra of titanium are very rich in lines which, in the visible region, were measured long ago by Thalen, 1 Angstrom, Cornu, 2 and Liveing and Dewar. 3 Lock- yer 4 records twenty-four additional lines between A, = 4000 and 3900, and Cornu twenty-five between A = 3510 and 3217. Hasselberg 5 has recently subjected the arc-spectrum, between D and A, = 3450, to a thorough investigation. For its production he introduced a fragment of rutile into the hollow of the carbon anode. Rowland has identified many of the solar lines with those of titanium ; they are not all given in the list below, but are included in his table of normal lines in Chapter IX. Arc-spectrum : 6261.3 6258.6 6222. o 6215. 2 6i26.3 6098. 6 6091. 6 6084.4^ 6065. Sg 5999-8 5979. 1 5966. 5 5953. 5922.7 5920.0 5899.56 5866.69 5804.45 5786.21 5774-27 5766.56 5762.52 5739-69 5715.30 5714.12 5702.92 5689.70 5680.15 5675-61 5662.37 5648.81 5644-37 5565.70 (5514.78 5514.58) 5512.72 1 N. A. S. U. (1868) [3] 6. 2 Spectre normal (Paris, 1881). Journ. de 1'Ecole polyt. (1883) 52. 3 P. R. S. (1881^ 32, 402. 4 P. T. (1881) 173, 561. 5 Svensk. Vetensk. Akad. Hand. (1895) 28, No. i. Astrophys. J. (1896) 3, 116; 4, 212. * Visible only in the spark-spectrum. (Thalen.) \ Visible only in the spark-spectrum. (Hartley and Adeney.) \ 3=98.9 cor. (Watts.) Visible in the spark-spectrum. 1" OF THB r UNIVERSITY SPECTRA OF THE ELEMENTS. 169 5504.10 5490.38 5488.44 (5482.09 5481.64) 5429-37 5409-81 5369.81 5297.42 5295.95 5283.63 5266.20 5256.01 5252.26 5238.77 5226.70 5225.15 (5224.74 5224.46 5223.80) 5219.88 5210.55 5193.15 5188.87 5I73.94 5152.36 5M7.63 5145.62 5120.60 5"3.64 5087.24 5064.82 5040.12 5038.55 (5036.65 5036.10) 5025.72 5025.00 5023.02 5020.17 5016.32 5014.49 5013-45 5007.42 4999.67 4997.26 4991.24 4989.33 4981.92 4975-52 4928.50 4921.90 4919.99 4913.76 4900.08 4885.25 4870.28 4868.44 4856.18 4841.00 4820.56 4799-95 4792.65 4778.44 4759-44 4758.30 4742.94 473L33 4723.32 4722.77 4710.34 4698.94 4693.83 4691-50 4682.08 4675.27 4667.76 4656.60 4650.16 4645-36 4640. 1 1 (4639-83 4639.50) 4629.47 4623.24 4617.41 4572 -15 4563-94 4555.64 4552.62 4549-79 4548.93 4544.83 4536.25 4536.12 4535-75 4534-97 4534-15 4533-42 4527.48 4522.97 4518.18 4512.88 4501.43 4489-24 4481.41 4475.00 4471.40 4468.65 4465.96 4457-59 4455-48 4453-87 4453.48 4451.07 4449-32 4443-97 4440.49 4434- i 5 4430-55 4427.28 4426.24 4423.00 4417.88 4417.46 4404.42 4399-92 4395-17 4394-04 4369-82 4338.05 4326.50 4325.30 4321.82 4318.83 4314.95 4314.50 4313.01 4306.07 4302.08 4301.23 4300.73 4300.19 4299.79 4299.38 4298.82 4295.91 4294.28 4291.07 4290-37 4289.23 4287.55 4286.15 4285.15 4282.85 4281.49 4274.73 4263.28 4256.18 4238.00 4186 27 4171.15 4163.80 4I59.79 4151.11 4137.39 4127.67 4123.68 4112.86 4082.57 4078.61 4060.42 4055.18 4030.60 4026.64 4024.71 4021.98 4013.72 4009.06 3998.77 3989.92 3982.62 3981.91 3964.40 3962.98 3958.33 3956.45 3948.80 3947.90 3930.02 3926.48 3924.67 39 J 4.45 39I3-58 3304.95 3901.13 3900.68 3895-42 3883.02 3882.49 3882.28 3875.44 3873-40 3869.47 3868.56 3866.60 3862.98 3858.26 3853.87 3853.18 3822.16 3786.20 3771.80 3761.46 3759.42 3753.75 3753-00 374LI9 3729.92 3725.28 3724.70 3722.70 3694-58 3690,04 3685.30 3671.82 3669.08 3662.37 3660.75 3659-9 1 3658.22 3654.72 3653.61 3646.32 3642.82 3641.48 3635.61 3610.29 3599-25 3598.87 3596.17 3547-15 3535.56 3530.53 3510.98 3505-02 349 I -20 3477-33 TUNGSTEN. The spectrum of this element is obtained by means of sparks from a Leyden jar, and, in the visible region, has been measured by Thalen. 1 In the arc-spectrum Lockyer 2 observed seven additional lines between A, = 4000 and 3900. N. A. S. U. (1868) [3] 6. 2 P. T. (1881) 173, 561. I7O SPECTRUM ANALYSIS. Spark-spectrum: 5734.1 5514.1 5492.6 5224.2 5071.4 5068.9 5053.9 5014.9 5007.9 4888.5 4843.1 4302.6 4295.6 4269.6 URANIUM. Thalen 1 measured the lines produced by the passage of powerful sparks from a Leyden jar through a solution of the chloride, and Lockyer 2 observed fifty-six lines in the arc- spectrum between A 4000 and 3900. The uranium salts exhibit characteristic absorption-spectra, which have been investigated by H. W. Vogel, 3 Morton and Bolton, 4 and Zimmennann; 5 some of them are shown in the following chapter. The spectra of the uranic and uranous salts differ; the latter consists of a strong double band in the orange, a feebler band in the green, and a broader one in the blue; these are shown when a uranic salt is reduced by means of zinc and hydrochloric acid, and are not affected by the pres- ence of iron, chromium, cobalt, nickel, zinc, or aluminium (Vogel). Spark-spectrum : 5914.1 5620.1 5580.2 5563-7 5523.1 55io.i 5494.6 5482.5 5480.5 5478.0 5475.5 5385.1 5027.9 4732.0 4724.0 4543-9 4473.4 4394-3 4374-7 4362.7 4341.2 VANADIUM. Thalen 6 has measured the spark-spectrum, and Lockyer has mapped fifty lines in the arc-spectrum between A. = 4000 and 3900, whilst the region between \ = 4450 and 1 N. A. S. U. (1868) [3] 6. 2 P. R. S. 27, 280. P. T. (1881) 173, 561. 3 Prakt. Spectralanalyse (Berlin, 1889). 4 C. N. (1873) 28, 47, 113, 164, 233, 244, 257, 268. 5 Zeitschr. Anal. Chem. 23, 221. See also Oeffinger, Inaug.-Diss. Tubingen, 1866. Stokes, P. A. Suppl.-Bd. (1854) 4, 273. Hagenbach, P. A. (1872) 146, 395. 6 N. A. S. U. (1868) [3] 6. 1 P. R. S. 27, 280. P. T. (1881) 173, 561. SPECTRA OF THE ELEMENTS. 171 4030 has been examined by Hasselberg 1 in his investigation of the occurrence of vanadium in rutile. Arc and spark spectra: 6241.7 6120.1 6090.2 6040.2 5726.1 5703.6 5698.6 5669.1 5627.1 5623.6 5415.4 5241.1 5234.2 4881.9 4875.5 4844. i 4594.30 4586.55 4580.59 4577o6 4545.81 4469-87 4462.53 ' 4460.39 4459.92 4452.17 4444.40 4441.90 4438.01 4416.65 4408.65 4408.40 4407.90 4406.85 4400.75 4 395- 40 4390.15 4384.90 4379-42 4353.05 434LI5 4333- co 4330.15 4271.80 4268.85 4134.60 4128.25 4123.65 4116.65 4H5.65 4115-30 4112.00 4109.95 4105.30 4100.00 4095.60 4092.87 4090. 70 YTTERBIUM. The spark-spectrum of the ytterbia earths has been meas- ured by Thalen; 2 no absorption-spectrum of the salts is known. Spark-spectrum; 6221.9 6005.0 5984.4 5837.0 5819.0 5556.6 &476.9 5353.0 5347-4 5345-9 5335.0 4936.0 4786.5 4725.9 YTTRIUM. The spectrum measured by Thalen 3 was obtained by passing powerful sparks through a solution of the chloride. Lockyer 4 observed twenty-six lines in the arc-spectrum between A. = 4000 and 3900. Rowland 5 considers it pos- sible that yttrium is composed of two substances, he includes the following ultra-violet lines in his table of wave- lengths; they were observed by means of the arc: 3950.50, 3774.48, 3710,44, 3633-28, 3611.20, 3602.06, 3600.88, 3584.66, 3549.15. Crookes ' has described a phosphorescent spectrum of yttria. 1 Bishang till Svenska. Vetensk. Akad. Handl. (1897) 22 Afd. I. No.7 ; 23. Afd. I. No. 3. Astrophys. Jour. (1897) 5, 194 ; 6, 22. 8 Ofvers. K. Vetensk. Akad. Forhandl. 1881. * Om spectra Yttrium, Erbium, Didym Och Lanthan, (Stockholm, 1874), 4 P. T. (1881) 173, 561. 8 Astronomy and Astrophysics (1893), 12, 321. Johns Hopkins Univ. Circulars (1894), 13, 73. See also Bunsen, P. A. 155, 366. 6 P. T. 174, 891. A. c. p. [6] 3, 145. P. R. S. 35, 262: 38, 414. 172 SPECTRUM ANALYSIS. Phosphorescent spectrum of yttria: Bands: 6676.7 6180.7 5737.9 5492-4 540O-5 4825.7 4323-6 Spark-spectrum : 6614.0 6435.5 6191.4 6181.9 6164.5 6150.1 6131.9 6019.5 6009.5 6003.5 5987.4 5663.0 5605.6 5577-1 5545-6 5544-1 5527.5 5521.0 5510.0 5497.0 5480.4 54739 5466.9 5403.0 5206.0 5200.5 5123.3 5118.8 5088.3 4881.9 4855.0 4643.8 4527.3 4422.7 4374.6 4309.6 4177.1 4167.7 ZINC. The spectrum of zinc has been measured by Kirchhoff, 1 Huggins, 2 Thalen, 3 Lecoq de Boisbaudran, Mascart, 4 Cornu, 6 Lockyer, 6 Liveing and Dewar, 7 Hartley and Adeney, 8 Ames, 9 and Kayser and Runge. 10 The last observers volatilized the metal or chloride in the carbon arc. The spark-spectrum differs considerably from that of the arc, probably on account of the higher temperature of the former. Spark and arc spectra: 6363.7* 6103.0* 4924.6* 4912.0 4812.2 4810.71 4722.26 4680.38 4630-06 3345-62 3345.13 3303.03 3302.67 3075.99 3072.19 3035.93 3018.50 2801.00 2771.05 2756-53 2712.60 2684.29 2670.67 2608.65 2582.57 2567.99 2558.03 2542.53 2516.00 2491.67 2246.90 2099- if 2073.7! 2061.3! 2024.6! 3282.42 2770.94 2570.00 2138.3$ 1 A. B. A. 1861. 2 P. T. (1866) 154, 139. 3 N. A. S. U. (1868) [3] 6. 4 Ann. de 1'Ecole normale (1866), 4. 6 J. de phys. (1881) 10, 425. C. r. (1885) 100, 1181. P. T. (1873) 163, 253, 639. 1 P. R. S. 29, 402. P. T. (1883) 174, 205. 8 P. T. 175, 63. P. M. (1890 V[ 5 ] 30, 33- 10 A. B. A. 1891. * Only in the spark-spectrum (Thalen). f Only in the spark-spectrum (Corun). \ Only in the spark-spectrum (Ames). SPECTRA OF THE ELEMENTS. 1 73 ZIRCONIUM. The visible portion of the spark-spectrum has been meas- ured by Thalen, 1 who passed powerful sparks through the chloride. Lockyer 2 observed twenty-three lines in the arc- spectrum between A = 4000 and 3900. Spark-spectrum : 6344.8 6311.3 6141.8 6133.7 6128.1 5350.5 5191.7 4816.1 4772.1 4739.5 4710.5 4687.5 4155.7 4149.7 1 N. A. S. U."(i868) [3] 6. 2 P. T. (1881) 173, 561. See also Troost and Hautefeuille, C. r. (1871) 73, 620. CHAPTER VIII. ABSORPTION-SPECTRA. AN attempt to treat absorption-spectra in the same systematic manner as emission-spectra is attended with con- siderable difficulty. The latter have in many cases been thoroughly investigated, and all the lines measured on a uniform system of wave-lengths, but observations of the former are almost exclusively confined to the visible region, although some investigations of the ultra-violet and infra-red regions have given very promising results. Many of the observations have been made by the help of instruments unprovided with measuring appliances, or, when these were present, only arbitrary scales were employed. The majority of observers omit details of the concentration of the solutions, and of the thickness of the layer of liquid, so that the results plotted as curves are not comparable with one another. Kriiss' method of determining the minimum of brightness is perhaps well adapted to give results meeting the above conditions, but hitherto it has not been generally employed. Some of the problems which have arisen during the later development of spectroscopic work should be capable of solution by the study of absorption-spectra, but this can only occur if the investi- gations are conducted on a totally different plan, and the greater portion of the work hitherto accomplished must be repeated. 1 1 Comp. Hasselberg, K. Svenska. Vetensk. Akad. Handl. (1891) 24. No. 3. 174 ABSORPTION-SPECTRA. 1/5 These criticisms are not intended to imply that the study of absorption-spectra has been entirely lacking in important results; the weighty and thorough investigations of the solar spectrum which belongs to this class of spectra, and is treated separately in the following chapter, would of themselves be sufficient to prove the contrary. In addition numerous im- portant observations have been made in various other direc- tions; some of those referring to inorganic substances arc mentioned in the preceding chapter. In the majority of instances, although the results have some value as preliminary observations, they lack the degree of accuracy attainable at the present time, and are of such unequal value that they can scarcely be included in a work such as this, where detailed criticism would be out of place, whilst the redrawing of curves on a uniform scale instead of the arbitrary ones, would introduce fresh sources of confusion and error. The account given in the following pages will be restricted to a record of the laws deducible from the results of observa- tions; further details can be obtained by consulting the original memoirs, or H. W. Vogel's " Praktische spectral- analyse, " Berlin, 1889, which contains numerous illustrations of absorption-spectra. To show the method usually adopted for the graphic re- cording of the results, a table of some well-known absorption- spectra is given in Fig. 41. The spectra of inorganic com- pounds include those of salts of chromium, copper, cobalt, iron, manganese, nickel, and uranium, whilst the organic compounds are alizarin, aniline blue, chlorophyll, eosin, fluorescei'n, fuchsin (magenta), indigo, malachite green, methyl violet, purpurin, quinoline red, safranine, and blood in aqueous solution, both alone, and after treatment with reagents; these are useful for the identification of blood- stains, and for the diagnosis of carbon monoxide poisoning. Absorption by Gases and Liquids. The laws of the absorption by gases have been discussed in Chapter VI. The . SPECTRUM ANALYSIS. TABLE OF ABSORPTION-SPECTRA. C D Eb F G 760 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 8 Chrome alum(violet) inwater. Chrome alum(green) inwater. Ferric chloride in water. Ferric thiocyanate in ether. Cobalt chloride (solid). Cobalt chloride in alcohol. Cobalt chloride in water. Cupric sulphate. Potassium permanganate in water. Potassium manganate in water. Nickel chloride in water. Uranium nitrate in water. Uranous salt m water. > Alizarin in sulphuric acid. Aniline blue soluble in water. Blood. Oxyhaemoglobin. Blood. Oxyhsemoglobm, re- duced. Blood. Haematin in hot al- cohol. Blood. Hsematm in cold al- cohol. Blood. Hsematin in alcohol and ammonium sulphide. Blood. Carbon monoxide haemoglobin. Quinoline red in alcohol. Chlorophyll(fresh) in alcohol. Eosin in alcohol. Fluorescein in alcohol. Fuchsin in alcohol. Indigo in sulphuric acid or chloroform. Malachite green. Methyl violet. Purpurin in sulphuric acid. Safranine salt with i equiv. of acid. Safranine salt with 2 equivs, of acid. B Eb FIG. 41. ABSORPTION-SPECTRA. 1 77 absorption by liquids differs, as Ostwald 1 has shown, accord- ing to whether or no the substance is a salt. In the case of colored bodies other than salts, the color is a constitutive property, as is well exhibited in the case of organic dyes. 2 Every change in the molecule produces a definite alteration in the absorptive power, so that variation in the absorp- tion is correlated with change in the molecular condition (Stenger). The absorptive power of salts in dilute solution is purely additive, and is the resultant of the sum of the color of the ions. v Different Salts of the same Colored Base or Acid. The first systematic investigations of the absorption-spectra of different salts of the same acid were made by Gladstone; 3 his results show that, in general, a base or acid retains its absorptive power in its compounds. Arrhenius' electrolytic dissoci tion theory indicates that aqueous salt solutions become more dissociated as the dilution increases, until a limit is reached at which the properties become wholly addi- tive, and are the sum of those of the ions. Ostwald has tested this theory by an examination of the color of salt solu- tions. The absorption-spectrum of a salt, in a solution of infinite concentration, would be rather complex, as the solu- tion would contain at least three constituents, viz., undecom- posed molecules, and the two ions; the spectrum would therefore be compounded of three distinct absorption-spectra, the relative intensity of which would be proportional to the quantity of each constituent present, and the undecomposed molecules might exist in aggregations of varying degrees of complexity. With this property, as with all others, increas- 1 Ueber die Farbe der lonen. Abhandl. K. sach. Ges. d. Wissensch. 31. Zeitschr. phys. Chem. (1889) 3, 601 ; (1892) 9, 226, 579. Compare also Mag- nanini, Zeitschr. phys. Chem. 12, 56. O. Knoblauch, W. A. (1891) 43, 738. Wagner, Zeitschr. phys. Chem. 12, 314. Ewan, P. R, S. (1895) 57, 117. 2 Stenger, W. A. (1888) 33, 577- 3 P. M. (1857) 14, 418. 178 SPECTRUM ANALYSIS. ing dilution causes simplification as the molecular aggregates are resolved, until a limit is attained at which the phenomena are caused solely by the ions. The spectrum of a highly dilute salt solution is therefore the sum of the spectra of its ions. The matter may be further simplified by selecting compounds with one colorless ion, i.e., one that exerts no absorptive power on the particular region of wave-lengths under examination. Hence it follows that in highly dilute solutions compounds of a particular colored ion with any colorless ones must give identical absorption-spectra; this is in accordance with Kundt's rule, and apparent exceptions appear to be all due to secondary influences. Ostwald has photographed the absorption-spectra of about 300 salts, including those of the following substances: permanganic acid, fluorescein, eosin, iodeosin, tetrabromorcinolphthalei'n, dinitrofluorescei'n, rosolic acid, diazoresorcinol, diazoresorufin (resorufin), chromium oxalates, safranines, rosanilines, aniline violet, chrysaniline, and chrysoidin. All th^ compounds investigated conformed with the above rule; apparent excep- tions were observed in the case of certain feeble acids or bases, the salts of which are hydrolyzed, but the addition of excess of the base or acid overcame this divergence, and the behavior of the solution was then normal. Some salts are insoluble, and therefore not dissociated; these also are appar- ent exceptions. Occasionally the salt forms visible deposits of a colloidal nature, but if the dilution prevents this, a determination of the electrolytic conductivity will immediately establish the absence of dissociation. Relationship between the Molecular Structure and the Absorption-spectrum. Many investigations have been made to determine the influence of chemical composition on the position of the absorption-bands; a number of regularities have been observed, particularly in the case of organic com- pounds, but no law of general applicability has been formu- lated. ABSORPTION-SPECTRA. 179 Absorption in the Visible Portion of the Spectrum. The primary object of the earlier workers was the general investigation of the spectra, rather than their relationship to the chemical constitution of the compounds. The chrysoidins were the first dyes, the color of which could be changed at will from pale yellow to red by the systematic introduction of various groups; these, and other azo-dyes of known constitu- tion, were examined by Landauer; 1 the spectra obtained were not very sharp, but it was shown that the absorption-bands undergo a marked change when the hydrogen in the amido- group is replaced by methyl. G. Kriiss investigated indigo 2 and fluorescem 3 derivatives, and arrived at the following con- clusions. The substitution of hydrogen by methyl, ethyl, methoxyl, carboxyl, or any group which increases the per- centage of carbon in the compound, causes the absorption- bands to approach the red; the effect of bromine is similar. The minima of brightness approximate towards the blue if hydrogen is replaced by the nitro- or amido-groups; the dis- placement increases as the number of substituting groups rises, and, if the groups are identical, it is proportional to their number. Kruss' rule, except the proportional displace- ment, was confirmed by Liebermann and Kostanecki 4 in the case of methylated hydroxyanthraquinones, and by Bernthsen and Goske 5 in that of dimethyl- and diethylthionine. The latter workers observed that the introduction of two methyl groups produces a displacement of about 2Oyw//, the change caused by two additional methyl groups being about 45/^u. E. Vogel 6 subsequently proved that the proportional displace- ment of the absorption-bands is the exception instead of the 1 Ber. (1881) 14, 391. 2 G. Kriiss and S. Oeconomides, Ibid. (1883) 16, 2051. 3 Ibid. (1885) 18, 1426. * Ibid. (1886) 19, 2327. 5 Ibid. (1887) 20, 924. See also Bernthsen, Studium in der Methylene- fclaugruppe. Lieb. Ann. (1885) 230, 73. Goske, Inaug.-Diss. Zurich, 1887. 6 W. A. (1891) 43, 449. 180 SPECTRUM ANALYSIS. rule; in the eosin compounds the displacement depends both on the number and position cf the substituting groups. In aqueous solution the displacement varies according to whether the substituting group enters the phthalic acid, or the resor- cinol nucleus. Influence is also exercised by the position, in the phthalic acid radicle, of the substituting group, if any, and also by the. solvent employed. H. W. Vogel 1 had previously shown that the extent to which the absorption- bands are displaced is intimately related to the position of the substituting group in the molecule. He investigated methy- lated azo-dyes in concentrated sulphuric acid solution. Grebe 8 worked on the same subject, and examined more than a hundred azo-dyes; his conclusions are as follows: In sul- phuric acid solution the absorption-bands of the azo-dyes approach the red as the content of carbon increases. Hy- droxyl- and amido-groups act in a similar manner; constancy in the position of the radicles produces a constant displace- ment of the bands; thus if hydroxyl is introduced into the naphthalene molecule, the ^-compounds exhibit bands which are always about 2O//yu nearer to the red than those of the isomeric ytf-derivatives. The sulphonic group produces a dis- placement in the opposite direction; it is almost constant, and approximates to 4O/^w; the bands are more clearly defined, and the influence of the position of the group is also nearly constant. In connection with the relationship between color and absorption-spectrum Schutze's 3 theory of dyes may be men- tioned. A study of physics shows that the mixture of colors composing the solar spectrum appears white because each color has an equally strong complementary color; these neutralize one another, arid produce on the eye the effect of whiteness. If a color is abstracted from the spectrum by 1 Sitzungsber. Berl. Akad. (1887) 34, 715. Ber. 21, 7760. 2 Zeitschr. f. phys. Chem. (1893) 10, 673. Ber. (1893) 26, 1300. 3 Zeitschr. f. phys. Chem. (1892) 9, 109. Meyer's Jahrb. 1892, p. n. A BSORP TION-SPE CTRA . 1 8 1 absorption, the issuing light will be tinged with the comple- mentary color. Many colorless substances exhibit absorption- bands in the ultra-violet; the introduction of groups which displace the bands towards the red (" catho-chromic groups ") first produces absorption in the violet, and the substance con- sequently assumes a greenish yellow color. " Hypsochromic groups " displace the absorption-bands towards the violet, and their introduction causes the opposite effect ; such groups are not numerous. * This explanation is in accord with the em- pirical rule, enunciated in 1877 by Nietzki, which states that the simplest dyes are greenish yellow or yellow; as the molec- ular weight increases, the color changes successively to orange, red, violet, blue, and green. In the case of analogous elements a rise in the atomic weight also produces an increase in the depth of the color; an example is furnished by the fluorine group: this element is light greenish yellow, chlorine is a deeper greenish yellow, bromine red, and iodine violet. Experience shows that the color of many organic compounds is dependent on the presence in the molecule of certain groups, such as the azo-group in the azo-dyes; this suggests that these complexes are probably the active agents in producing the absorption of light by the molecule, and that change in the color caused by substitution is due to influence exerted by the substituting radicles on these groups. 1 This view was first developed by O. N. Witt, a who applied the term 41 chromophore " to the group producing the color ;^since the influence of a substituting group on the chromophore will be stronger the nearer their positions approximate in the mole- cule, the theory affords a prospect of determining the relative distance of the groups by spectroscopic measurement. v Schutze examined certain azo-dyes, and found that the relative dis- tance of the atoms in the molecule indicated by the structural 1 Compare Hartley, J. Chem. Socy. (1887) 51, 152. Ber. (1887) 20, 131. 2 Ibid. (1876) 9, 522. 1*2 SPECTRUM ANALYSIS. formulae of the compounds corresponded, on the whole, with the spectroscopic measurements. Spring, 1 using extremely thick layers, observed that "colorless" organic compounds exhibit spectra free from absorption-bands if their molecules consist of carbon chains about which the heterologous atoms or groups are equally or symmetrically divided. Concentration of the atoms or radicles at one end of the chain causes the appearance of absorption- bands. The number of the bands appears to be directly related to the number of the hydrocarbon radicles present in the compound; their position is apparently determined by the nature of each group, but if two of these are very intimately linked, the combined influence causes a change in the position of the corresponding bands, which may even merge into a single one. Absorption in the Ultra-violet. The relationship be- tween the constitution of substances and their absorption- spectra in the ultra-violet region was first shown by Soret and Rilliet, 2 and Hartley and Huntington. 3 The former examined the ethylic, isobutylic, and amylic salts of nitric and nitrous acids; these compounds are not well suited to investigations of this nature, as tne absorption is confined to one end of the spectrum, and does not include measurable bands, but their observations rendered it probable that the whole absorption- field is displaced towards the red if a . hydrogen atom is replaced by methyl. Hartley and Huntington 3 obtained more definite results, showing that the normal fatty acids have a stronger absorptive power for the refractive rays of the ultra-violet region than the corresponding alcohols, and that increase in absorptive power in this part of the spectrum is correlated with increase in the number of CH 2 groups in the molecule of the homologous alcohols and acids. 1 Bull. Acad. roy Belgique [3] 33, 165. Chem. Centralbl. (1897) 68, 1114. 2 C. r. (1879) 89, 747. 3 P. R. S. (1879) p. 233. P. T. (1879) 170, 257. A BSORP TION- SPECTRA . 183 Hartley, 1 partly in conjunction with Huntington, 8 has carried out further extensive investigations of the relationship between the absorption-spectra of carbon compounds and their molecular structure. The following is a summary of the results: The alcohols CnH 2n + 2 , ethers, and ethereal salts (esters) readily transmit the ultra-violet rays; methyl alcohol almost as completely as water; but the fatty acids absorb these rays to a greater extent than do the corresponding alcohols. None of the compounds examined exhibit absorption-bands. In the case of the alcohols and acids the absorption increases as the content of carbon in the compound rises. Benzene and its hydroxyl-, carboxyl-, and amido-derivatives have a high absorptive power for the ultra-violet rays, and in thin layers exhibit strong absorption-bands. The spectra of isomers differ both in the position of the bands and in the extent of absorption. In the cresols and dihydroxybenzenes the meta-derivative has the greatest, and the para-compound the least absorptive power, but orthoxylene and parahy- droxybenzoic acid absorb more light than the isomeric compounds. The spectra of isomeric terpenes differ; the compounds C 10 H 16 and C JB H 24 exceed benzene in absorptive power, the former to a greater extent than the latter. Ab- sorption-bands are confined to compounds containing ben- zenoid carbon linkages. A simple linkage of carbon and nitrogen is insufficient to produce characteristic absorption of the ultra-violet rays. The substitution of a nitrogen atom for a CH group in benzene or naphthalene derivatives (pyridine or quinoline compounds) does not destroy the selective absorption-power, but this disappears if the benzenoid linkage is destroyed by combination with hydrogen. The addition of four atoms of hydrogen to carbon atoms in the quinoline 1 J. Chem. Socy. (1882); (1885) 47, 685; (1887) 51, 58; 53, 641. Ber. (1885) 18, 592; (1887) 20, 174; (1888) 21, 689. 2 P. R. S. 28, 233; (1879) 29 > 290; (1880) 31, i. P. T. (1879) 170, 257. Ber. (1881) 14, 501. 1 84 SPECTRUM ANALYSIS. molecule only diminishes the intensity of the absorption- bands. Molecules consisting of dissimilar parts vibrate as wholes. The fundamental vibrations produce secondary, ones which do not bear any recognizable relationship to the chemical constituents. Hartley has also examined the carbohydrates and albumi- noids, and has investigated the physical connection of these compounds with the soluble ferments. The spectra of egg albumin, serum albumin, and casein exhibit certain bands in common which are absent in the spectra of malt diastase, yeast invertase, gelatin, starch, glycoses, and saccharose, solutions of which are particularly transparent to the violet and ultra-violet rays. The albumins are thus shown to differ considerably from the ferments in constitution, and this accords with the difference in behavior shown by the com- pounds towards carbohydrates. Absorption in the Infra-red. The influence of the atomic grouping of organic compounds on their absorption of the infra-red rays has been extensively investigated by Abney and Festing. 1 Hydrogen chloride shows a few lines, water lines and bands; ammonia, and nitric acid sharp lines. Many of these lines coincide, and can be referred to hydrogen. It is to some extent an open question why the hydrogen acts as if it were free, and produces a line-spectrum, but certain absorption-lines exhibited by hydrocarbons are coincident with some shown by compounds of hyrdogen and oxygen, or hydrogen and nitrogen, which are known to be due to hy- drogen. The behavior of oxygen varies according to whether it is in the radicle; if it links radicles, it produces a continuous absorption between two hydrogen lines. Oxygen in the radicle increases the sharpness of the bands, and causes them to be bordered by lines. 1 P. R. S. 31, 416; (1881) 32, 258. ABSORPTION-SPECTRA. 185 Organic radicles are characterized by well-marked band? which chiefly occur between about 700^ and looo/*/*; some have a distinct absorption at about 7OOjw/*, and a second in the neighborhood of 900;^, but the characteristic absorption of the radicle is almost always in the former region. The ethyl group has an absorption at 742 /f/u, and a characteristic band between 892///t and 92O/*/u. The characteristic line of the phenyl group is at 867/*//. A comparison of the spectra of ammonia, benzene, aniline, and dimethylaniline shows a very close coincidence, and proves how slight an effect is pro- duced by varying the mass of the radicles linked to the nitrogen atom. Abney and Festing were unable to demon- strate the presence of haloids in organic compounds by means of their absorption-spectra in the infra-red. CHAPTER IX. THE SOLAR SPECTRUM. The Fraunhofer Lines The solar spectrum is the most complex of all absorption-spectra. It is permeated with a number of vertical lines which were first observed by Wollas- ton 1 in 1802, and investigated by Fraunhofer, 8 after whom they are named. He prepared a drawing of the spectrum, which is reproduced in Fig. 42, and contains about 350 lines. The more prominent are designated by the Latin letters A to H. By means of accurate measurements Fraunhofer showed that the. lines are constant both in relative position in the spectrum, and also in mutual distance. He observed that the spectra of the moon and planets are identical with that of the sun, but differ from those of Sirius and other fixed stars, so that the lines must originate in the sun and fixed stars, and not in the earth's atmosphere. In 1824 he proved the coincidence of the double ZMine of the solar spectrum with that of the sodium-spectrum, but was unable to explain the origin of the dark lines. Nine years later Brewsfer 1 found that when the sun is low, so that the light traverses a considerable thickness of atmosphere, new lines become visi- ble; these are produced by atmospheric absorption, and were first accurately mapped by Brewster 4 alone, and in conjunc- tion with Gladstone. 5 Angstrom and others almost succeeded 1 P. T. 1802, p. 365. 2 Denkschr. d. Miinchener Akad. 1814-15. P. T. E. 1833. 4 P. M. [3] 8, 384. P. A. (1836) 38, 50. 5 P. T. (1860) 150, 149 1 86 THE SOLAR SPECTRUM. I8 7 Red Orange Yellow Green Blue Indigo Violet FIG. 42. 1 88 SPECTRUM ANALYSIS. in showing the origin of the lines, and this was finally accom- plished by Kirchhoff 1 in 1859, as a consequence of the law of exchanges which he had recently formulated (comp, Chapter VI). The explanation of their origin offered a prospect of determining the composition of the sun and fixed stars, so that they at once received considerable attention. It became necessary, therefore, to measure the position of the Fraunhofer lines, and also those in the spectra of the elements, with a greater degree of accuracy than had been hitherto attained; this was done by Kirchhoff. 2 who used a prism spectroscope, and consequently an arbitrary scale. Angstrom, in the prep- aration of his atlas of the solar spectrum which appeared in 1868, made the measurements in wave-lengths, and, like Fraunhofer, employed diffraction gratings. His normal spec- trum included about 1000 lines, all based upon wave-lengths, and the atlas remained the foundation of all wave-length determinations for more than twenty years. The accuracy of his work was scarcely exceeded by the more extensive charts of H. C. Vogel and Muller, 3 Fievez, 4 Piazzi-Smyth, 6 Thollon, 6 and Young. Mascart, 7 Draper, 8 and Cornu 9 have investigated the ultra-violet region of the solar spectrum; the scale of the chart and the wave-lengths of the last observer correspond with those of Angstrom. Abney 10 photographed the infra-red region, and used his own special plates, which are sensitive to Monatsber. Berl. Akad. V. 27, Oct. 1859. A. B. A. 1861. Publicationen des Astrophys. Obs. zu Potsdam (1879), 1. Annales de 1'Observatoire Royal de Bruxelles (1882) [3], 4; (1883) 5. T. R. S. E. (1880) 29, 285; 32, 37, 233, 519. C. r. 88, 80. Ann. de 1'Observ. de Nice (1890), 3. Ann. de 1'Ecole normale sup. (1864) 1. Sillim. Jour. 1873. P. A. (1874) 151. C. r. 86, 101, 315, 530. Ann. de 1'Ecole norm. [2] 3, 421; 9, 21. Spectre normal du soleil (Paris, 1881). 10 P. M. [5] 6, 154; 11, 300. P. T. (1880) 171, 653: (1886) 177. W. A. Beibl. 4, 375; 5, 507, 509. Abney and Testing, P. R. S. 35, 80,324. W. A. Beibl. 8, 507. THE SOLAR SPECTRUM. 189 these rays. The spectrum was at first obtained by means of prisms, but subsequently with a Rowland's concave grating. Lommel 1 has also photographed the infra-red Fraunhofer lines, and Langley 3 has investigated them by means of the bolometer (comp. Chapter V). Mascart and Cornu extended Fraunhofer's designation of the lines by Latin capitals up to U '= 2948 Angstroms, and Abney introduced the following signs for the infra-red: ^ I 27000 Y\ W, \ Z70C * \ 8986.5 X, 8497.0 Calculated on $t 12400 X, 8806.1 z g2 | Angstrom's scale. #! 12000 X 3 8661.4 J After Angstrom's death it was found that his determina- tions were about ^lAnr to sma ll in consequence of his having used an incorrect metre-scale, so that new and more exact wave-length measurements became necessary. Muller and Kempf 3 made careful investigation of 300 solar lines, and almost simultaneously Rowland, in 1886, published the first edition of his photographic atlas of the normal solar spectrum. A superior and more complete edition 4 appeared in 1889. It consists of ten tables, each measuring 3X2 feet, and con- taining two spectrum sections. It includes the region between 3000 and 6950 Angstroms, and the scale which is attached is o correct to within less than 0.05 Angstrom. Photographs which are made by the sun itself are obviously superior to charts prepared by drawing and measurements. As an appen- dix to the atlas, Rowland 6 has published a table of the wave- lengths of numerous solar and metallic lines, the results of ten years' observations; they have been determined by his coinci- 1 Comp. Chap. V. 2 Sillim. Journ. [3] 25; 28; 31; (1886) 32; (1888) 36; (1889) 38. W. A. (1883) 19, 226, 384; (1884) 22. Beibl. 9, 335. * Publicat. d. Astrophys. Obs. zu Potsdam (1886) 5. 4 Photographic map of the normal solar spectrum. Baltimore, Md., Johns Hopkins Press. The price of the tables is $2.50, or $20 the set. 5 Astronomy and Astrophysics (1893), 12, 321. P. M. (1894) [5] 36, 49. 1 90 SPE CTR UM A NA L YSIS. o dence method, and are exact within o.oi Angstrom; in many o instances the accuracy is within o.ooi Angstrom, i.e., I in 5,ooo,ooo. 1 The value 2) l -= 5896.156 is taken by Rowland as the basis of his tables (comp. Chapter VII). An exact knowledge of the position of the solar lines is of the greatest importance for the orientation of spectra, and for the determination of the constituents of the sun; therefore in the following table Rowland's measurements of the Fraun- hofer lines are reproduced in full. The letter preceding a line is the one in general use for its designation; chemical symbols following it indicate the element with a line of which it coincides. A query (?) after the symbol means that it is doubtful whether the line does belong to the spectrum of the element. Two symbols after a line, such as 3295.957 Mn-Di, implies that both elements have a coincident line of this wave- length. Two or more symbols in brackets show that a line of the first element corresponds with one side of the solar line, one of the second with its middle, etc. Lines which do not coincide with those of any known element are followed by a query, whilst those due to absorption by atmospheric oxygen or moisture are designated by atm. (O) or atm. (H 2 O). All the wave-lengths are given in Angstrom units, reduced to air at 20 C., under a pressure of 760 mm. The Chemical Composition of the Sun. The establish- ment of KirchhofTs law, and the experimental reversal of the sodium lines (comp. Chapter VI) naturally directed attention to the elucidation of the composition of the sun. KirchhofTs careful drawings of spectra were prepared for the purpose of comparing the position of solar and metal lines. He employed the Fraunhofer lines for the orientation of metallic lines, and found that the iron lines coincided exactly with similar dark lines. That the coincidence of only sixty lines should be a 1 A list of the lines in Rowland's tables has appeared in the Astrophys. Jour, from 1895 onwards. THE SOLAR SPECTRUM. 19! ROWLAND'S TABLE OF WAVE-LENGTHS OF THE FRAUNHOFER LINES. Based on >i = 5876.156 Angstroms, and reduced to air at 20 and 760 mm. pressure. | 3005.160 3005.404 p p 3134.223 3I37.44I Ni Co 3377.667 Ti) Tif 3012.557 p 3140.869 Fe 3389.887 Fe 3014.274 3016.296 j Fe 3J53.870 3158.988 Fe Ca 3405.272 Co) Ti f 3024.475 ? 3167.290 Mn 3406.581 Fe 3025.394 p 3172.175 Fe? 3406.955 Fe 3025.958 Fe 3176.104 La? 3425.721 ? 3035.850 ) 3188.164 Cr? 3427.282 Fe 3037.492 Fe 3195.702 Ni 3440.759 Fe 3044.119 Ca[?Co] 3200.032 Ti "O] 3441. 135 Fe 3044.683 Mn 3214.152 Fe 3444.032 Fe 3046.778 p 3218.390 Ti 3455.384 Co 3047.720 Fe 3219.697 Fe 3464.609 Sr? 3050.212 3053-I73 3053-527 j Fe ? 3219.909 3222.203 Fe F e ??[ 3465.991 3475.594 Co) Fef Fe 3055-821 p 3224.368 Ti 3476.831 Fe 3057.557 Fe 3225.923 Fe Co) 3059.200 Fe Ti) 3478.001 Fe V 3061.098 ? 3231.421 ? I Nij 3061.930 Co 3232.404 Ti 3486 036 Ni 3067.363 Fe 3236.697 Ti 3490.721 Fe 3075.339 3075.849 Ti Fe 3246.124 u 3491.464 c \ 3077-216 Fe 3247.680 Cu 3497.264 Fe 3077-303 3078.148 p Fe? 3260.384 Mn ) Ti 3497-991 t\ 3078.759 Ti Fe ) 3500.721 Fe 3079.724 Mn 3267.839 V 3500.993 Ni 3080.863 p 3274.092 Cu 3510.987 Ti 3082.272 Al 3287.791 Ti 3513.947 Fe 3083.849 3086.891 Fe p 3292.174 Fe ) Co-T f 3518.487 3521.404 Co Fe 3088.137 Ti 3295.957 Mn-Di 3540.266 Fe 3092.824 Al 3302.501 Na 3545-333 p 3092.962 Al 3303-107 Na 3549-145 Yt 3094.739 3095.003 ? Fe 3303-648 4 3550.006 3558.670 Fe Fe 3100.064 3100.415 Fe Fe(Mn) 3306.117 3306.471 Fe Fe 3564.680 Ti) Fef 3100.779 3101.673 Fe Ni 3308.928 Mn ) Co-Ti f 3565.528 3570.225 Fe Fe 3101.994 Ni 3318.163 Ti 3570.402 Fe 3106.677 ? 333L74I Fe N] 3581. 344 Fe 3109.434 3115.160 Cr? Fe 3348.011 Cr) Fe f 3583-483 3584.662 Fe? Yt 3121.275 V 3351.877 Fe 3585.992 C 3129.882 Zr 3356.222 Zr 3586.041 C 1 9 2 SPECTRUM ANALYSIS. ROWLAND'S TABLE. Continued. 3590.523 3597-I9 2 C Fe 3736.969 Ni ) Mn ) 3883.773 3886.427 Cr Fe 3600.880 Yt(Fe) 3737-075 Ca 3897.599 Fe 3602.061 Yt 3737.282 Fe 3905.666 Si 3605.483 Cr Ti) 3916.875 Fe 3605.635 Fe 3743-502 Fe r 3924.669 Ti 3606.831 3609.015 Fe Fe 3745.701 Cr) Fe 3925-345 Fe) V f 3611.193 Yt 3746.054 Fe 3925.792 Fe 3612.217 Fe fa ) 3747-095 Fe ) 3926.123 F ? e ! 3617.92 *~"* f Fef 3748.409 Fe 3928.071 Fe 3618.924 Fe 3749.633 Fe KJ3933 809 Ca 3621.122 3621.606 Yt Fe 3754.664 l\ 3937-474 3941.021 Fe Fe-Co 3622.147 Fe 3756.2H Fe ? i 3623.332 Fe 3758.379 Fe 3942.559 Fe f 3623.603 Fe 3763.942 Fe 3944.159 Al 3628.853 Yt 3767.344 Fe 3949.C34 Ca 3631.619 Fe 3770.130 Fe 3950.101 Fe 3633.259 Yt 3774.480 Yt? 3950.497 Yt 3635-616 Ti 3780.846 ? 3954.001 Fe 3638.435 Fe 3781.330 Fe 3957.i8o Fe-Ca 3640.536 Cr ) 3783-674 3788.032 Ni Fe 3960.429 3961.676 Fe Al 3647.995 Fe 3794-014 Fe-Cr HJ3968.620 Ca 3652.692 Co 3795-150 Fe 397I-478 Fe 3653.639 Ti 3798.662 Fe 3973.835 Ca 3658.688 Mn) Fe f 3804.153 Fe Fe 3977.89I 3981.914 Fe Fe-Ti 3667.397 3680.064 Fe Fe 3805.487 [L] 3815.985 Fe-Di Fe 3984-078 Cr) Fe f 3683.202 Co) Fe[ 3820.567 3821.318 Fe Fe 3986.903 Mn [ v ) 3823.651 Mn(Cr) ? j 3683.622 Pb 3826.024 Fe 3987.216 Mn I 3684.259 Fe 3827.973 Fe Co ) 3687.607 Fe 3829.505 Mg 4003.916 Ce-Fe-Ti 3694-349 Yt 3832 446 Mg 4005.305 Fe-? 3695-194 Fe 3836.226 ?-C 4016.578 Fe 3705-711 3707.186 Fe Fe 3836.652 3838.430 C Mg 4029.796 Fe) Zr\ 3709.397 Fe 3840.584 Fe 4030.914 Mn 3710.438 Yt 3843.406 Fe 4033.225 Mn 3716.585 Fe 3856.517 Fe 4034-641 Mn 3720.086 Fe 3860.048 Fe . 4035.88 Mn 3722.691 Ni ) Fe-Ti f 3864.441 3871.528 C C 4044.293 4045.975 K Fe [M] 3 727-763 3732.542 Fe Fe 3875.224 vj- 4048.893 Mnl 3733-467 Fe 3883.472 C Cr n \ 3735.014 Fe 3883.548 C 4055.701 Mn THE SOLAR SPECTRUM. ROWLAND'S TABLE. Continued. T -93 4062.602 4063 756 Fe Fe Nil 4359-778 Cr \ 4691.581 Ti ) Fef 4071.904 Fe Zr) 4703.180 Mg 4073.920 Fe 4369.943 Fe 4703.986 Ni 4077.883 Sr 4376.103 Fe 4714.599 Ni 4083.767 Fe ) Mn f [d] 4383.721 Fe Fe ) 4722.349 Zn Fe ) 4083.928 Fe 4391.149 Tif 4727.628 Mn f 4088.716 Fe 4404.927 Fe 4754-226 Mn 4103.121 Si \ Mn | 4407-850 V ) Fe f 4783.601 Mn ? ) 4107.646 Fe 4413-181 Cd 4805.253 Tif 4114.600 Fe 4415.299 Fe 4810.723 Zn 4121.481 Cr) Co) 4425.609 4435.132 Ca Ca 4823.697 4824.325 Mn Fe? 4121.968 Fe-Cr 4435.852 Ca 4859.934 Fe 4157.948 Fe 4447.899 Fe T] 4861.496 H 4185.063 Fe 4454.950 Ca 4890.945 Fe 4197.251 C 4456.047 Ca 4900.098 Ti 4199.263 4202.188 4215.616 Zr) Fef Fe Fe) 4456.793 4494-735 4497.041 Ca Fe Cr) Zr f 4900. 306 4903-488 4919.183 Yt Cr) Fef Fe 4215-667 f 4499.070 Mn ) 4920.682 Fe 4215.687 Sr) 4499 315 ? \ 4924.109 Fe 4216.137 C 4501.444 Ti 4924.955 Fe 4222.381 Fe 4508.456 Ti? 4934.247 Ba [g] 4226 892 Ca 4554-213 Ba 4957.482 Fe 4250.290 Fe 4563.939 Ti 4957.786 Fe 4250.956 4254.502 Fe Cr 457L277 4572.157 # i 4973-274 TO Fcf 4260.638 4267.958 4271.924 Fe rU Fe 457,8.731 4588.384 4590.129 4602.183 Ca-Ti Cr? Ti? Fe 4978.782 4980.362 ? ) Fef "?} 4274 958 Cr 4607.509 Sr 4981.915 Ti 4283.170 4289.523 Ca Ca 46H.453 1 4994.316 4999.693 Fe Ti-La 4289.881 .4293.249 Cr ? 4629.515 SI 5005.904 5006.303 Fe Fe 4299.152 4302.689 Ca Ca 4637-683 4638.194 Fe Fe 5007.431 Ti ) Fef 4305.636 4306.071 Sr Ti 4643.645 4648.835 Fe Ni 5014.422 (Ni)Ti ) Ti r 4307.904 [G] 4308.034 4308.071 Ca l Fe} 4668.303 '4678.353 il Cd ' 5020.210 5036.113 Ti Ti^ Nif 4318.818 Ca 4679.028 Fe 5041.795 Ca [f] 4325 940 Fe 4680.319 Zn 5050.008 Fe 4343.387 Cr) Fej 4683.743 4686.395 Fe Ni 5060.252 5064.833 Fe Ti 4352.903 Fe 4690. 324 ? 5068.946 Fe I 9 4 SPECTRUM ANALYSIS. ROWLAND'S TABLE. Continued. 5083.525 Fe 5217.559 Fe 5383.576 Fe 5090.959 Fe 5225.690 Fe 5389.683 Fe 5097. 1 76 Fe 5230.014 Fe 5393.378 ' Fe 5105.719 Fe(Cu) 5233.124 Fe 5397.346 Fe 5109.825 Fe 5242.662 Fe 5405.987 Fe ? ) 5250.391 Fe 5410.000 Cr 5110.570 5115.558 Fe? Ni 5250.825 5253.649 Fe Fe Fe / 5415.421 y j- 5121.797 Ni ) FP \ 5260.557 Ca Ca ) 5424.284 | p e j- 5126.369 re j Co 5261.880 ^,0. 1 Crf 5434.742 Fe 5127.530 Fe 5262.341 ? I 5447.130 Fe Fe) 5262.391 Caj^ 5455-666 Fe? ) 5133-871 ? r 5264.327 Cr 5455-759 5139.437 Fe) 5264.371 f 5455-826 Fe ) 5139-539 I 5264.395 Ca) 5462.732 : Ni 5139.645 Fe ) 5265-727 Ca 1 5463.174 Fe 5141.916 Fe 5265.789 I 5463.493 Fe 5142.967 5143.042 Ni [ 5265.884 (Ni?) j Cr J 5466.608 Fe 5477.128 ; Ni 5143.106 Fe ) 5266.729 Fe 5487.968 Fe 5146.664 Nij [E 2 ] 5269.722 5270.448 Fe 5497-731 5501.685 Fe Fe 5151.026 Fe )- Mn f [E,] 5270.495 5270.533 Fe) 5507.000 5513.207 Fe Ca 5154-237 Ti?Co? 5273.344 Fe) 5528.636 Mg 5155.937 Ni 5273-443 \ 5535.073 Fe 5159.240 Fe? 5273-554 Fe) 5543.418 Fe 5162.448 Fe ? } 5544.158 Fe 5165.190 C 5276.205 Cr V 5555.113 Fe 5165.588 Fe Co ) 5569.848 Fe 5167.501 Mg) 5281.968 Fe 5576 319 Fe [b 4 ]5i67.572 V 5283.803 Fe 5582.195 Ca 5167.686 Fe ) 5288.708 Fe 5588.980 Ca 5169.066 Fe ) 5296.873 Cr 5590.342 Ca [b 3 ]5i69.i6i V 5300.918 Cr 5594.695 Ca 5169.218 Fe) 5307.546 Fe 5598.555 Fe 5171-783 Fe 5316.790 Fe?) 5598.715 Ca [b 2 ]5172.871 Mg [147415316.870 [ 5601.501 Ca 5173-912 Ti 5316.950 Co?) Fe) [b!]5183.792 Mg 5324-373 Fe 5603.097 Ca}. 5188.863 Til 5333.092 Fe? Fe) 5188.948 [ 5349.623 Ca 5615.526 Fe 5189.020 Ca) 5353-592 Fe-Ni 5615.879 Fe 5193.139 Ti 5361.813 ? 5624.253 Fe 5198.885 Fe 5363-011 Fe(Co) ) 5624.768 Fe-V p ) 5363.056 ? r 5634.167 Fe 5202.483 Fe \ 5367-670 Fe 5641.661 Fe 5204. 708 5210.556 Cr) Fef Ti 5370.165 5371.686 Fe Ni ) Fe-Cr f 5645.835 5655.707 5658.096 Si Fe Yt? 5215.352 Fe 5379.776 Fe 5662.745 Fe THE SOLAR SPECTRUM. ROWLAND'S TABLE. Continued. 195 5675.648 5679.249 Ti Fe 5930.410 5934.883 Fe Fe 6256.574 Ni) Fef 5682.861 Na 5948.761 Si 6261.316 Ti 5688.434 Na 5956.925 Fe 6265.347 Fe 5701.769 Fe 5975.576 Fe 6270.439 Fe 5708.620 Si 5977-005 Fe a] 6278.289 atm. (O) 5709.616 Fe 5977.254 atm. (H a O) 6281.374 atm. (O) 5709.760 Ni 5985.044 Fe 6289.608 atm. (O) 5711.318 Mg 5987.286 Fe 6293.152 atm. (O) 5715-309 Ni ) Fe-Ti f 6003.245 6008.196 Fe Fe 6296.144 6301.719 atm. (O) Fe 5731-973 Fe 6008.782 Fe 6314.874 Ni 574 2 -66 Fe 6013.717 Mn 6315-541 Fe 5752.257 Fe 6016.856 Mn 6318.242 Fe-(Ca) 5753-342 Fe A/~\OO 1A*7 ? ) 6322.912 Fe 5754.884 Ni| W4U. Jif/ 6022.017 Mn 6335.550 6337.042 Fe Fe 5763-215 Fe 6024.280 Fe 6344.370 Fe 5772.360 Si 6027.265 Fe 6355.259 Fe 5775.304 Fe 6042.316 Fe 6358.902 Fe 5782.346 Cu? Co? 6056.232 Fe 6378.461 Ni 5784-081 Cr 6065.708 Fe 6380.951 Fe 5788.136 Cr 6078.709 Fe 6393.818 Fe Cr) 6079.223 Fe 6400.200 ? 5791.207 Fef 6102.408 Fe 6400. 509 Fe 5798.087 ? Ca) 6408.231 Fe 5798.400 Fe 6102.941 Fef 6411.864 Fe 5805.448 Ni 6103.449 6420.171 Fe 5806.954 Fe 6108.338 Ni 6421.569 Fe 5809.437 Fe 6111.287 Ni 6431.063 Fe 5816.594 Fe 6116.415 Fe 6439.298 Ca 5831.832 Ni 6122.428 Ca 6450.029 Ca 5853-903 5857-672 Ba Ca 6136.834 6141.934 Fe Fe-Ba 6462.835 Ca) Fef 5859.810 Fe 6154.431 Na 6471.881 Ca 5862.580 Fe 6160.970 Na 6480.264 atm. (H 2 O) [0315875.982 He 6162.383 Ca 6482.099 ? 5884.048 Fe ) at. (H,0) \ 6169.260 6169.775 Ca Ca 6494.001 6495.209 Ca Fe 5889.854 atm. (H 2 O) 6173-554 Fe 6499.871 Ca [D,]5890.182 Na 6177.028 Ni 6516 315 7 5893.098 Ni 6180.419 Fe 6518.594 Fe [D,]5896.154 Na 6191.397 Ni 6532.546 atm. (H 2 O) 5898.395 at. (H 2 0)) Fe? \ 6191.770 6200.533 Fe Fe 6534.173 Ti ) 5901.681 at. (H a O) j Fe? f 6213.646 6219.493 Fe Fe 6546.486 6552.840 Fef atm. (H 2 O) 5905.895 Fe 6230.946 Fe-V ;C] 6563. 054 H 59 T 4 384 Fe ) ?-at.(H,0)$ 6237.529 6246.530 7 Fe 6569.461 6572.312 Fe atm. (H 2 O) 5916.475 Fe 6252.776 Fe 6574.477 p 59I9-855 atm. (H a O) 6254.454 Fe 6575.179 Fe 1 9 o SPECTRUM ANALYSIS. ROWLAND'S TABLE. Continued. 6593.161 Fe 6889.194 atm. (O) 7148.427 ? 6594.115 Fe 6890. 149 atm. (O) 7168.191 p 6609.354 Fe 6892.614 atm. (O) 7176.347 at. (H 2 O)? 6633.992 Fe 6893-559 atm. (O) 7184.781 at. (H 2 O?) 6643.882 Ni 6896.292 atm. (O) 7186.552 at. (H 2 0^> 6663.525 p 6897.195 atm. (O) 7193.921 atm. (H.jO> 6663.696 Fe 6900 199 atm. (O) 7200 753 atm. ( H 2 O) 6678.232 Fe 6901.113 atm. (O) 7201.468 atm. (H.>()} 6703.813 p 6904.358 atm. (O) 7210.812 at. (H 2 6?> 6705-353 ? 6905.263 atm. (O) 7223.930 5 6717.934 Ca 6908.785 atm. (O) 7227.765 p 6722.095 p 6909.675 atm. (O) 7232.509 7 6726.923 Fe 69I3-454 atm. (O) 7233-171 p 6750.412 Fe 6914.328 atm. (O) 7240.972 atrn.(H.,O) 6752.962 Fe 6914.819 Ni 7243.904 aun.(HaO) 6768.044 Ni 6916.957 ? 7247.461 at. (H,Oj? 6772.565 Ni 6918.363 atm. (O) 7264.851 at. (H 2 Oi? 6787.137 Fe 6919.245 atm. (O) 7265.833 at. (H 2 O,? 6807.100 Fe 6923 557 atm. (O) 7270.205 p 6810.519 Fe 6924.420 atm. (O) 7273.256 at. (H 2 O)? 6820.614 Fe 6928.992 atm. (O) 7287.689 at. (H 2 O)? 6828.850 Fe 6929.838 atm. (O) 7290.714 at. (H 2 0)? 6841.591 Fe 6934.646 atm. (O) 7300.056 atm.(H 2 O) 6843.908 Fe 6935-530 atm. (O) 7304 475 at. (H 2 O)? 6855.425 Fe 6947.781 at. (H 2 0?) 7318. bi8 at. iH 2 Oj? 6867.461 6867 800 atm. (O) aim. (O) 6953-838 at. (H 2 O?) atm.? 7321.056 7331.206 p ) 6868.124 atm. (O) 6956.700 atm.(H 2 0) 7389.696 p 6868.393 atm. (O) 6959.708 at. (H 2 O?) 7409.554 p 6868.779 atm. (O) 6961.518 at. (H 2 O?) 7446.038 ? 6869.141 atm. (O) ) 6978.655 p 7462.609 p 6869.347 atm. (O) j" 6986.832 atm. (H 2 O) 7495o5i p [B] 6870.186 atm. (O) 6989.240 at. (H 2 0?) 7511.286 p 6871.179 atm. (O) 6999.174 at. (H 2 O?) 7545 921 ? 6871.527 atm. (O) 7000.143 ? r A1 j 7594.059 atm (O) 6872.493 atm. (O) 7006. 1 60 p LAJ { 7621.277 atm. (O) 6873.076 atm. (O) 70H.585 p 7623.526 atm. (O) 6874.039 atm. (O) 7016.279 at. (H 2 O?) 7624 853 aim. (O) 6874.884 atm. (O) 7016.690 at. (H 2 O?) 7627 232 aim. rO) 6875.826 atm. (O) 7023.225 p 7628.585 atm. (O) 6876.957 atm. (O) 7023.747 ? 7659 658 atm. (O) 6877.878 atm. (O) 7024.988 p 7660.778 Htm. (O) 6879.294 atm. (O) 7027.199 p 7665 265 atm. (O) ( 880.176 atm. (O) 7027.726 ? 7666.239 atm. (O> 6881.970 Cr 7035.159 ? 7670.993 atm. (O) 6882.772 Cr 7038.470 ? 7671.994 atm. (O> 6883.318 Cr 7040.058 ? 7699.374 p 6884.083 atm. (O) 7090.645 ? 7714.686 p 6886.008 atm. (O) 7122.491 p 6886.987 atm. (O) 7M7 942 ? THE SOLAR SPECTRUM. 1 97 matter of accident is practically out of the question, as he * calculated that the chance of this being so is only I in a trillion 3 -line, produced by the ele- ment which Frankland termed helium, and to the bright green line of the corona designated by Kirchhoff 1474^. As alieady stated, Ramsay 2 has recently obtained helium from cleveite. Limits of the Investigation. It is somewhat surprising that so many terrestrial elements, such as the non-metals, and the metals of high atomic weight, appear to be absent from the sun, but the investigation can only proceed a certain length. It has been already stated that, in the ultra-violet, the solar spectrum does not extend beyond about 3OO/*yu. As the temperature rises spectra tend to develop in the violet; hence, on account of the extremely high temperature of the sun, a considerable portion of its spectrum must neces- sarily escape observation. Cornu 3 states that the absorption of the ultra-violet region is not caused by the varying con- stituents of the atmosphere, such as water vapor or dust, but essentially by nitrogen and oxygen. He 4 has suggested a formula for the calculation of the length of the solar spectrum absorbed by the column of air which the light traverses; according to this, a thickness of 663 metres causes o a diminution of 10 A. at the ultra-violet end. The formula indicates that the extreme limit which can be observed is w.-l. = 2930, and it also shows that at w.-l. = 2120 and w.-l. 1570 total absorption is caused by strata of air 10 m. and o.i m. in thickness, respectively. This was confirmed by experiment: the triple line of aluminium of w.-l. = 1860 was rendered unrecognizable by passage through a column of 1 C. r. (i860) 63, 289, 728. A. c. p. (1871) [4] 23, 274 ; 24, 215. 2 C. N. (1893) 71, 151. 3 C. r. 90 940 4 ''ibid. 88, 128= ; C9, 808. THE SOLAR SPECTRUM. 2O3 air 4 m. in length. A further obstacle to the determination of the chemical composition of the sun is the fact that its nucleus, comprising at least nine tenths of the whole, is not available for spectroscopic investigation, as explained below. The Physical Condition of the Sun. In order to explain the occurrence of the dark lines in the solar spectrum Kirch- huff concluded that the atmosphere of the sun encloses a luminous mass which emits a continuous spectrum of high illuminating power. This inner portion is either solid or liquid, and at a higher temperature than the atmosphere. Subsequent investigations, both under ordinary conditions and during solar eclipses, have shown that the sun is more complex than Kirchhoff imagined. A complete treatment of the subject is altogether beyond the scope of this work, par- ticularly as opinion is still much divided; the majority of investigators agree with C. A. Young's ' views, and it will suffice to attempt a brief sketch of these. The nature of the inner nucleus of the sun can only be conjectured, as it is beyond the reach of observation. Probably it consists of gas at an extremely high temperature, and under such an enormous pressure that its properties must resemble, to some extent, those of a viscous substance like putty. Surrounding the nucleus is the photosphere, composed of glowing cloud-like masses of vapor; it forms the visible sur- face, and appears to correspond with the clouds in the terres- trial atmosphere. It is not known whether it is separated from the nucleus by a definite surface; externally, it is sharply but irregularly defined, being elevated in some places into faculce, and in others depressed, forming spots. The reversing lavcr is situated directly over the photosphere, and produces the Fraunhofer lines; its thickness is only about 1000 miles. The gases composing the reversing layer are not confined ex- clusively to the surface of the photosphere; they also occupy 1 The Sun. New York and London, 1896. ^ UNIVERSITY 204 SPECTRUM ANALYSIS. the spaces between the photospheric clouds, and constitute the atmosphere, in which these float. Above the reversing layer is the scarlet-red chromosphere, consisting of uncon- densed gases (hydrogen and helium); from this numerous prominences extend far beyond the surface of the sun. The exterior portion of the sun is termed the corona; it consists of clouds and irregular streams of light, and gradually merges into the surrounding darkness. The greater portion of the mass of the sun is within the photosphere, but the larger part of its volume is outside it; the diameter of the solar atmos- phere is at least double that of the central portion, and its volume consequently seven times as great as this. The idea that the NUCLEUS of the sun consists of gas is supported by the fact that its atmosphere has a temperature sufficient to vaporize metals, and also because the sun's mean density is low. Compared with that of the earth it is only 0.253, or in comparison with water 1.406; it would necessarily be much greater than this if it consisted to a great extent of liquid iron, titanium, magnesium, etc. As the temperature of the gaseous mass is far ; bove its critical point, the high pressure must cause it to exceed water in density, and there- fore the gases must be viscous, and comparable in properties with molten glass or putty. The PHOTOSPHERE is undoubtedly a gaseous envelope, condensed in places to cloudlike masses of vapor in conse- quence of the heat radiating into space. Its irregular appear- ance is due to these masses, the solid or liquid particles of which cause its luminosity, and produce a continuous spec- trum like the solid particles in an ordinary flame. The spectrum of the SUN-SPOTS exhibits a number of dark bands: the dark lines of calcium, iron, titanium, etc., are widened, and some lines, like those of hydrogen, are often reversed; the sodium lines are also frequently enormously widened, and doubly reversed, as shown in Fig. 43. These phenomena render it likely that the increased absorption is due to gases THE SOLAR SPECl^RUM. 2O$ % and vapors rushing in to fill a space, and absorbing the light emitted from the cavity. In consequence of the violent motion of the gases, lines are some- times displaced, as explained in the following chapter. The FACUL^E show the H and K bands of calcium, always reversed by a thin bright line running down the middle of each ; and, whilst the re- versal directly over a spot is generally FlG - " single," it is usually " double " in the faculous region sur- rounding it, i.e., the bright line is double. This makes it somewhat probable that the faculae are not mere protrusions from the photosphere, but luminous masses of calcium vapor floating in the solar atmosphere, and possibly identical with the prominences themselves. The emission spectrum of the REVERSING LAYER can only be observed during a total eclipse; at the moment when the sun is completely obscured by the moon the lines of the whole spectrum are seen to flash out brightly luminous. Like the other phenomena, the spectra of the CHROMO- SPHERE and its PROMINENCES were formerly only visible during an eclipse; but in 1868 Janssen ' and Lockyer a inde- pendently, and almost simultaneously, devised a method by which these portions of the sun could be observed daily in a clear atmosphere. Zollner* and Huggins 1 have suggested similar methods of procedure. A spectroscope of high dis- persive power is employed, and the slit opened widely; if not too large, the whole prominence is then visible. The promi- nences appear to bear a certain relationship to the sun-spots and faculae; they are divided into two classes quiescent, 1 C. r. (1868) 68, 93. 2 P. R. S. (1868) 17, 91, 104. 128. s P. A. (1869) 138, 32. 4 P. R. S. (1868), 17, 302. 200 SPECTRUM ANALYSIS. cloudlike, or hydrogen and helium prominences, and erup- tive or metallic ones. The former resemble terrestrial clouds in appearance; the latter are highly luminous, but the degree of luminosity and the shape change with extreme rapidity. Their spectra is very complicated, and, as shown by the dis- placement of the lines, they often attain a velocity exceeding 100 miles per second. The size of the prominences varies between wide limits; the mean thickness of the chromosphere is about 7500 to 9500 km. (5000 to 6000 miles), therefore no prominence can be less than 7000 to 9000 miles. Secchi observed 2767 prominences; of these 1964 attained a height of 29,000 km. (18,000 miles), and 751 exceeded 43,000 km. (28,000 miles). Young, in 1880, observed a prominence extending a distance of 562,400 km. (350,000 miles), the longest hitherto noticed. The following lines form the spectrum of the true chromosphere, and are -always present: 1. 7065,50 He 7. O] 4340.66 \\y 2. [C] 6563.05 H 3 ] 5875-98 He 9. 3970.20 He 4. [1474 A'] 5316.87 Coronal line 10. [H] 3968.56 Ca 5. [A] 4861. 50 H/J ii. [AT] 3933.86 Ca 6. [/] 4471.80 He There are numerous additional lines sometimes visible; their occurrence depends on the comparative violence of motion of the soJar atmosphere. The principal ones are the following: 6678.2 He [ 3 ] 5169.16 Fe 4491-5 Mn 6431.1 Fe [ 4 J 5167.57 Mg 4490.2 Mn 6141.93 Ba 5018.6 Fe 4469.5 Fe [Z>i] 5896.2 Na 5015-7 He 4245.5 Fe [Z? 2 j 5890.2 Na 4934-25 Ba 4236.1 Fe 5363 Fe? 4924.11 Fe 4233.8 Fe 5284.6 Ti ? 4922.2 He 4226.9 Ca 5276.21 Cr? 4919.1 Fe? 4215.67 Sr 5234.7 Mn 4900.3 Ba 4077.88 Ca 5198-2 ? 4584.1 Fe 4026.0 He [<*i] 5183-79 Mg 4501.44 Ti 3889.1 H [ 2 J 5172.87 Mg THE SOLAR SPECTRUM. 2O/ The Corona. The corona is visible only during a total eclipse, and much uncertainty prevails as to its nature. In 1869 Harkness, Pickering, and Young independently observed the coronal line 1474 AT., of wave-length 5316.87, to which reference has already been made. Of the element which pro- duces this line nothing is at present known; that it is far less dense than hydrogen is probable from the fact that the line remains clear and sharp during the most violent movements of the prominences. It was long believed that this bright line was the only one present in the spectrum of the corona, but others, including those of hydrogen and calcium, have been subsequently observed, chiefly by Schuster, 1 who in 1882, in Egypt, was able to detect about thirty. In addition to the bright lines Pickering and Eastman in 1869 noticed a faint continuous spectrum, in which Janssen, and also Barker detected some of the stronger Fraunhofer lines, D, b, G. It is now generally admitted that the corona consists of an atmosphere extending 300,000 miles, and of extreme tenuity. The nature of the streamers is still uncertain; some regard them as a sort of permanent aurora, their position and direc- tion being determined by the sun's magnetic field of force, as the terrestrial fields of force direct the beams of the aurora borealis. Schaeberle believes them to be due to light emitted and reflected from streams of matter ejected from the sun by forces acting, in general, along lines normal to the surface of the sun, and most active near the centre of each sun-spot zone. The attempts of Huggins 2 and others to photograph the corona in ordinary daylight have not been successful. 1 Abney and Schuster, P. T. 1884. 8 P. R. S. 34, 409; 39, 108. Astron. Nachr. 104, 113. W. A. Beibl. 9, 755- CHAPTER X. OTHER CELESTIAL BODIES. 1 AURORA BOREALIS. ZO- DIACAL LIGHT. LIGHTNING. DISPLACEMENT' OF THE LINES. Fixed Stars. The physical condition of the fixed stars resembles that of the sun. Their continuous spectra, traversed by rectangular dark lines, shows that they consist of an incandescent mass surrounded by a glowing atmosphere. Fraunhofer, in 1817, observed that the dark lines differ in the spectra of various stars, and that they do not correspond with those in the solar spectrum ; various stellar spectra were also correctly characterized by him. After the foundation of an accurate system of spectrum analysis Secchi and H. C. Vogel divided stellar spectra into groups, according to the degree of development of the stars. The arrangement is as follows: Class I. Stars at such high temperatures that the metallic vapors in their atmospheres exhibit only a slight absorptive power (white stars). The spectra consist (a) of strong hydrogen lines and feeble metallic lines (Sirius, Vega, and the majority of white stars); (b) of single feeble metallic lines without the strong hydrogen lines (ft, y, tf, e Orionis); (c) of bright hydrogen lines, and the bright _/9 3 -line (<* Lyrae, y Cassiopeiae). Class II. Stars which resemble the sun, have an atmos- phere containing metals, and exhibit a spectrum containing strong absorption-lines. The spectra show, (a) in addition to 1 Comp. Scheiner, Die Spectralanalyse der Gestirne (Leipzig, 1890), where a detailed account is given of the subject, with references to the literature. 208 OTHER CELESTIAL BODIES. 2OQ the hydrogen lines, numerous strong metallic lines, especially in the yellow and green (Capella, Arcturus, Aldebaran); (b) numerous bright lines together with dark ones, and a few faint bands (T^Coronae). Class III. Stars at such a low temperature that the sub- stances composing their atmospheres have combined to form chemical compounds which produce absorption-bands (red stars). The spectra exhibit, (a) in addition to dark lines, bands, dark and sharply defined towards the violet, whilst towards the red they become irregular (OL Herculis, OL Orionis, ft Pegasi). Most of the lines are due to iron, but it is undetermined whether the bands consist of aggregates of fine lines, or of strong lines extended laterally; (b) dark, very broad bands, sharply bounded towards the red, and gradually disappearing towards the violet. This type only includes stars of small mangitude, which is unfortunate, as the spectra suggests the possibility of their atmospheres containing glow- ing carbon. The Planets and Moon. Since the planets and moon reflect sunlight, their spectra must be essentially that of the sun, modified by absorption-lines or bands produced by their own atmospheres. Spectroscopic observation shows that Mercury, Venus, and Mars have atmospheres similar in nature to that of the earth, and containing aqueous vapor. The same applies to Jupiter and Saturn, but their spectra exhibit an additional absorption-band; whether this is caused by differences in temperature and pressure or by the presence of a new gas is undetermined. The atmospheres of Uranus and Neptune also differ materially from that of the earth, and contain an additional constituent in large quantity. The spectrum of the moon is in every way identical with that of the sun, showing that it has no atmosphere, or only one of extreme tenuity. The satellites of Jupiter exhibit the same spectra as that of the planet itself, and appear to have identical atmospheres. 210 SPECTRUM ANALYSIS. Comets. The first spectroscopic observation of a comet was made by Donati in 1864; he found that the spectrum consisted of three bright bands superposed on a continuous spectrum, and that, in part at least, the comet was self- luminous. Four years later Huggins stated that the bright bands were identical with those obtained by passing electric sparks through ethylene. Subsequent accurate measurements made by H. C. Vogel, and Hasselberg showed that this is not the case, although the comets undoubtedly contain carbon in considerable quantity. A spectrum very similar to that of a comet is obtained by passing a continuous electric discharge through a mixture of a h\drocarbon and carbon monoxide. The comet I of 1882 was observed by H. C. Vogel to contain sodium; this was confirmed by Duner and Bredichin, whilst Copeland and J. G. Lohse noticed iron lines in comet II of 1882, which passed within a few thousand miles of the sun. Huggins, in 1881, observed that the continuous spectrum exhibits the Fraunhofer lines, proving that a portion at least of its light is reflected sunlight. Photometric and spectro- scopic observations of a sudden outburst of light in the case of comet I of 1884 showed that a part of the continuous spectrum is due to the comet's own luminosity. Hasselberg has suggested that this is probably caused by electrical forces, and observations of other kinds have rendered it very probable that comets are the seats of electrical activity. Meteors and Shooting Stars do not lend themselves to spectroscopic observation on account of the short period during which they are visible. They exhibit a continuous spectrum, caused by the incandescence of the solid constit- uents, but in addition to this only the sodium line has been observed with certainty. The meteorites which fall on to the earth can all be analyzed by the ordinary chemical methods, so that their spectroscopic investigation is not of much im- portance. Nebulae. The spectroscopic investigation of the nebulae OTHER CELESTIAL BODIES. 211 is of considerable interest in connection with the Kant- Laplace hypothesis of the origin of the solar system. Formerly they were classified as divisible nebulae, which could be resolved into clusters of stars, and indivisible or true nebulae; but the latter were regarded as being capable of resolution if sufficiently powerful telescopes were available. The first spectroscopic observations of nebulae were made by Huggins in 1864; he noticed the existence of bright lines, showing the presence of luminous gases. The faintness of the light renders the investigation a matter of difficulty; with medium instru- ments three or four lines only are usually visible. The wave- lengths are about 5004, 4957, 4861, 4341. By means of photographic processes about forty additional lines may be detected. The presence of hydrogen is known with certainty, and that of helium is probable, but the origin of the majority of the lines is unknown. Many of the nebulae exhibit a faint continuous spectrum in addition to the bright lines; its maxi- mum is in the green, instead of the yellow. Vogel states that it shows no sign of discontinuity, but Copeland and Huggins consider that it appears to be resolved into lines. If this view is correct, the nebulous clusters of stars are masses of glowing gas, and are to be regarded as stellar systems, the individuals of which are gaseous. Aurora Borealis. This phenomenon is the result of elec- tric discharges in highly rarefied air, and has been a frequent subject of spectroscopic observation. The spectrum exhibits a number of very faint lines, together with a characteristic bright green one of wave-length = 5571, which has been termed the aurora line. Its origin is not known, but H. C. Vogel, Zollner, and Hasselberg agree in regarding the remain- ing portion of the spectrum as a modification of that of air. The spectrum of the ZODIACAL LIGHT is a reflected solar spectrum; the faintness of the light necessitates the use of a wide slit, so that the Fraunhofer lines are unrecognizable. The spectrum of LIGHTNING has been examined by many, 212 SPECTRUM ANALYSIS. including Kundt, John Herschel, Laborde, H. C. Vogel, Joule, Procter, Young, and Schuster. The majority have detected the line spectrum of nitrogen, frequently in combina- tion with a continuous spectrum, and occasionally with a band-spectrum of unknown origin. Schuster states that this last bears a very close resemblance to the spectrum, at the cathode, of a vacuum tube containing oxygen mixed with a small proportion of carbon monoxide. DISPLACEMENT OF THE LINES. The spectroscope has rendered important help to astrono- mers in elucidating the relative velocity of bodies in the line of collimation. In the sections on sun-spots and prominences it was mentioned that their extremely rapid motion produced a displacement of the spectrum lines. The explanation of this phenomenon is obtained from Doppler's 1 principle, first propounded in 1841, according to which the color of the light received on to the retina, or the pitch of a note changes if the source of light or sound approaches or recedes from the observer at a speed not too small in comparison to that of light or sound respectively. If the source of light or sound approaches the observer more waves will be received in a given time than if it were stationary, whilst if it is receding the number of waves will be less. V The color or wave-length of a ray from an object approaching will therefore be diverted towards the violet, but will approximate to the red if the object is receding. yThe alteration, to a stationary observer, according to whether the light approaches or recedes, is given by the expression X, = A 1 1 -], where A = the wave-length of the ray, A, = that produced by the motion, v being the velocity of light, and a that of the luminous body. 1 Ueber das farbige Licht der Doppelsterne und anderer Gestirne des Himmels. Abhandl. K. Bohmischen Ges. d. Wissensch. (1841-2) [5] 2, 465. OTHER CELESTIAL BODIES. 213 44 shows the displacement of the /Mine in the spectrum of a sun-spot. Huggins, in 1864, first employed the displacement of the lines to determine the velocity of Sirius in the line of FIG. 44 collimation. He observed that the wave-length of the had increased O.iOQyU//; the velocity of light 297,100 km. per second, and the wave-length of the /<-line = 486.5^. 297100 X 0.109 Consequently the expression Q^; ~ 66.6 shows that, at the time of the observation, Sirius and the earth were receding at the rate of 66.6 km. per second ; but the earth was itself moving from Sirius at the rate of 19.3 km., so that the speed of the latter is reduced to 47.3 km. Subsequent observations, with improved instruments, led Huggins to- modify this to 29 35 km. per second. Similar investiga- tions of the stars and nebulae have been made by H. C. Vogel, Seabroke, and the Greenwich astronomers. Lockyer and Young have employed the method for the determination of the velocity of portions of the solar atmosphere, so far as they move in the line of collimation. The speed with which- changes take place is enormous, and often resembles a violent cyclone; the rising and sinking masses of gas in the spots, attain a velocity of 30 to 50 miles per second, whilst that of the prominences is frequently 150 km. (100 miles) per second, and occasionally twice as gre.at. INDEX OF AUTHORS. Abney, 7, 62, 151, 184, 188, 207 Adeney, 8, 97, 101, 104, 108, 121, 131, 135, 142, 143, 145, 148, 151, 154, 160, 165, 166, 167, 172 Allen, 108 Ames, 20, 45, 88, 89, 108, 128, 164, o I72 Angstrom, 4, 6, 62, 76, 92, 104, 109, no, 114, 117, 118, 128, 135, 144, 146, 148, 149, 151, 153, 154, 164, 188, 197, 201 Attfield, 107, in, 162 B Bahr, 124, 125 Bailey, 124 Balmer, 8, 80, 129 Becker, 151, 201 Becquerel, E.,62, 63, 96, 104, 143, 157, 162 Becquerel, H., 63, 109, 124 Beilstein, 155 Bell, 92, 108, 152 Bernthsen, 179 Berthelot, 99, 128 Boisbaudran, Lecoq de, Q, 58, 65, 80, 86, 87, 88, 96, 97, 102, 105, 108, 10^, ii?, 117, 118, 121, 122, 124, 125, 126, 130, 131, 135, 142, 143, I44 145, 146, 151, 152, 154, 155, 156, T 57 J58, 161, 162, 163, 166, 167, 172 Bolton, 170 Brauner, 127 Brewster, 2, 118, 144, 151, 152, 186 Briihl, 150 Brunner, 136 Bunsen, 5, 20, 58, 65, 79, 95, 102, 108, 116, 122, 124, 125, 141, 142, 143, 157, 158, 162, 163, 166, 171 C Cappel. 96 Capron, 8, 130 Cauchy, 37 Cazin, 8 Chappuis, 151 Christiansen, 16 Christie, 20, 31 Christofle, 155 Ciamician, 89, 101, 103, 105, 115, 117, 132, 156, 160, 164, 166 Clayden, 131 Cleve, 127, 141, 158 Clifton, 5 Collie, 127 Cornu, 7, 59, 6r, 77, 81, 96, 103, 108. 109, 119, 128, 135, 143, 144, 146, 151, 162, 166, 168, 172, 188, 201, 202 Crookes, 98, 103, 124, 166, 171 Currie, 63 D Daniell, 105. 132 Delachanal, 57 Delafontaine, 122 y, 8 215 i6 INDEX OF AUTHORS. Desains, 61, 63 Deslandres, 59, 69, 81, 86, 114, 115, 126, 130, 148, 149, 154 Dewar, 8, 54, 77, 81, 96, 97, 102, 104, 108, 109, no, in, 114, 117, 119, 121, 125, 126, 130, 131, 135, 141, 142, 143, 144, 145, 146, 150, 154, 157. J58, 160, 161, 162, 166, 167", 168, 172, 199 Diacon, 53, 122, 142, 157, 162 Dibbits, 8, in, 114, 152 Dieterici, 9 Ditte, 89, 101, 117, 164, 165 Doppler, 212 Dorn, 99 Dove, 149 Draper, 9, 62, 115, 198 Dupr6, 58 E Ebert, 52, 68 Eder, 53, 68, 98, 99, 105, 109, no, ii2, 145, 152, 160, 163 Egoroff, 151, 201 Eisig, 154 Emsmann, 31 Erdmann, 99 Erhard, 118 Etard, 118, 119 Ewan, 121, 130, 136, 177 Ferry, 62 Festing, 184, 188 Fievez, in, 143, 151, 188, 199 Foucault, 4, 76 Frankland, 73 Franz, 61 Fraser, 153 Friedlander, 100 Fraunhofer, 4, 17, 186 Fuchs, 31 G Gange, 9 Galitzin, 73 Gernez, 132, 152, 160, 164, 165 Gladstone, 64, 122, 144, 145, 151, 177, 186 Glan, 40, 122 Glazebrook, 20 Goldstein, 151 Goske, 179 Gouy, 52, 155 Grace, 130 . Gramont, A. de, 9, 157, 160, 162,. 164 Grandeau, 9 Grebe, 180 Greiner, 136 Grubb, 25 Griinvvald, 89 H Hagenbach, 170 Haitinger, 124 Hale, 127 Hartley, 8, 57, 89, 97, 101, 103, 104, 105, 108, 112, 121, 131, 135, 136, 142, 143, 144, 145, 148, 150, 154, 155, 160, 165, 166, 167, 172, 181, 182, 183 Hasselberg, 59, 67, 97, 105, 117, 119, 129, 132, 146, 148, 149, 152, 154^ 164, 168, 171, 175 Hautefeuille, 89, 105, 112, 151, 160, 173 Helmholtz, H. v., 16, 60 Helmholtz, R. v., 62 Hennesay, 151 Herschel, Alex., 31 Herschel, A. S., 114, 115 Herschel, John, 2 Herschel, William, 61 Heycock, 131 Higgs, 9 Hittorf, 71, 77, 101, 105, m, 115, 117, 128, 132, 141, 148, 151, 153, 155, 160, 163 Hofmann, 152, 155 Holden, 156 Hiifner, 40 Huggins, 6, 9, 10, 97, 101, 102, 104, 109, in, 117, 118, 124, 126, 127,. INDEX OF AUTHORS. 217 128, 130, 135, 142, 144, 145, 151, 153. J55, 156, 157, 161, 162, 165, 166, 167, 172, 205, 207 Humphreys, 128 Huntington, 101, 182, 183 Hutchins, 154, 156 Janssen, 28, 130, 151, 154, 2OI, 2O2, 205 Jewell, Lewis E., 154 Johnson, 108 Julius, 62, 69, 87, 96 K Kahlbaum, 2 Kayser, 8, 9, 49, 54, 67, 69, 81, 87, 88, 89, 96, 97, 99, 101, 102, 104, 107, 108, 109, in, 114, i2t, 126, 127, 129, 131, 135, 141, 143, 145, 157, 158, 161, 162, 1 66, 167, 172, 199 Kelvin, 10 Kempf, 7, 92, 189 Kessler, 31 Ketteler, 142 Kirchhoff, 2, 5, 6, 20, 76, 95, 96, 97, 101, 102, 103, 108, 109, 116, 117, 118, 121, 122, 126, 134, 135, 141, 142, 143, 145, 146, 151, 155, 156, 157, 158, 161, 162, 167, 172, 188, 197 Klinkerfues, 9 Knoblauch, O., 79, 177 Kobb, 125 v. Kovesligethy, 9, 87 Kohn, 136 v. Konkoly, 9 Kopp, 2 v. Kostanecki, 179 Kruss, G., 9, 38, 64, 65, 118, 122, 124, 126, 144, 159, 179 Kriiss, H., 9, 64, 118, 122, 144 Kundt, 16, 78 Kuppelwieser, 136 Kurlbaum, 92 Lamansky, 61 Landauer, 64, 177 Langley, 7, 60, 62, 63, 151, 189 Lapraik, 118, 130 Lassel, 10 Lewis, 62 v. Lichtenfels, 136 Liebermann, C., 179 Lielegg, 9, 112, 136 Lippich, 73, 124 Listing, 95 Littrow, 25 Liveing, 8, 54, 77, 81, 96, 97, 102, 104, 108, 109, in, 114, 117, 119, 121, 125, 126, 130, 131, 135, 141, 142, 143, 144, 145, 146, 150, 154, 157, 158, 160, 161, 162, 166, 167, 168, 172, 199 Lockyer, 7, 9, 69, 73, 77, 97, 98, ior y 1 02, 103, 108, 109, no, in, 115, 116, 117, 119, 122, 127, 128, 134, 135, 141, 144, 146, 153, 155, 156, 157, 161, 162, 164, 166, 167, 168, 169, 170, 171, 172, 173, 198, 205 Lommel, 10, 63, 189 Lorscheid, 9 M MacMunn, 9 Magnanini, 152, 177 Mascart, 20, 104, 108, 109, 135, i6r, 162, 167, 172, 188 Masson, 4 Melloni, 61 Melville, Thomas, 2 Mermet, 57 Miller, W. A., 3, 7, 105, 132, 146, 166, Mitscherlich, 5, 53, 102, 104, 109, 115, 122, 125, 132, 141, 152, 160, 163 Monckhoven, 59 Morghen, 132 Morren, in, 115, 117, 152 Morton, 170 Moser, 105, 152 Mouton, 61 Miiller, 7, 92, 118, 136, 142, 162, 188, 189 218 INDEX OF A UTHORS. Miiller, J., 61 Miiller C-Pouillet), 10 Mulder, 155, 160, 164 N Neovius, 154 Newall, 99 Newton, 15 Nilson, 124, 159 O Oeconomides, S., 179 Oeffinger, 170 Orstnan, 119 Ostwald, 44, 177 Paalzow, 154 Palmieri, 127 Parker, Spear, 136 Paschen, 62, 127, 128, 154 Pearce, 145 Peirce, 92 Pfaundler, 10 Piazzi-Smyth, 59, III, 114, 130, 148, 151, 154, 188 Pickering, 20, 129 Pliicker, 5, 58, 71, 101, 105, in, 114, 115, 117, 128, 132, 141, 145, 148, 151, 153. 155, 160, 163 Pringsheim, 77 Proctor, 9 Ramsay, 98, 127, 202 Rayleigh, 16, 20, 98 Reich, 131 Reusch, 20 Reynolds, 80, 118, 124 Richard, 128 Richards, 100, 106, 117, 132, 149 Richter, 131 Rilliet, 182 Roberts, 98, 141. 166, 167 Roscoe, 5, 9. 72, 79, 105, 136 Rosen berger, 2 Rowland, 7, 20, 45, 92, 103, 105, 126, 153, 154, 156, 157, 158, 160, 166, 171, 189, 190, 199, 2OI Rubens, 60, 62 Riihlmann, 142 Runge, 8, 20, 54, 67, 69, 81, 87, 89, 96, 97, lor, 102, 104, 107, 108, 109, in, 114, 121, 126, 127, 128, 131, 135, 141, 142, 143, 145, 154, 157, 158, 161, 162, i(>6, 167, 172, 199 Russell, 119, 130 Rutherfurd, 26, 157, 162 Rydberg, 8, 69, 84, 109, 128, 162 Sabatier, 118, 122 Sabine, 121 Salet, 10, 53, 58^ 59, 105, 112, 117, 125, 128, 132, 148, 154, 155, 157, 159, 160, 164, 165, 167 Sarasin, 130 Scheiner, 10, 44, 208 Schellen, 10 Schonn, 130 Schottlander, 124 Schumann, 60, 61 Schiitze, 180 Schuster, 8, 69, 72, 80, 124, 148, 152, 153, 164, 207 Seabroke, 129 Secchi, 10, 135, 151 Seguin, 125, 155, 163 Sellmeier, 16 Sieben, 16 Siemens, W. v., 77 Snow, 62 Sorby, 40 Soret, 60, 80, 124, 130. 158, 182 Spring, 182 Stenger, 78, 177 Stewart, Balfour, 76 Stokes, 2, 6r, 76, 170 Stoney, 80, 87, 118 Swan, 3, in, 162 INDEX OF AUTHORS. 2I 9 Tait, 10 Talbot, Fox, 2, 114 Tatnall, 103, 105, 126, 153, 154, 156, 157, 158 Thal6n, 7, 10, 86, 92, 96, 97, 101, 102, 103, 104, 105, 108, 109, no, in, 114, 115, 116, 117, 118, 121, 122, 124, 126, 131, 132, 135, 141, 142, 143, 144, 145, 146, 148, 149, 150, 153, 154, 155, 156, 157, 158, 159, 161, 162, 165, 166, 167, 168, 169, 170, 171, 172, 197 Thiele, 132 Thollon, 20, 25, 114, 151, 188 Thorpe, 105 Travers, 127 Troost, 89, 105, 112, 160, 173 Trowbridge, 100, 106, 117, 121, 132, 149. *54 Tuckerman, 10 Tunner, 136 Tyndall, 10 Valenta, 53, 98, 99, 105. 109, no, 145, 160, 163 Vierordt, 10, 39 Vogel, E., 179 Vogel, H. C., 129, 148, 149 Vogel, H. W., 10, 30, 44, 58, 64, 78, 117, 118, 119, 128, 132, 134, 135,136, 142, 144, 145, 146, 154, 167, 170, 175, 180, 199 W Wagner, 177 Walter, 44 Watts, 10, 93, in, 114, 115, 136, 145 Wedding, 136 Welsbach, Auer v., 124 Wesendonck, 112, 114, 115 Wheatstone, 4 Wiedemann, E., 55, 68, 69, 77 Witt, O. N., 181 Willigen, van der, 5, in, 117, 148, 149 Winkelmann, 10, 69, 81, 88 Wollaston, 3, 16, no, 137, 186 Wolf, 53, 142, 157, 162 Wolff, C.H., 119, 121 Wright, 10 Wiillner, 69, 71, 73, 112, 114, 115, 129, 132, 148, 154 Young, 10, 203 Zeeman, 75 Zimniermann, 170 Zollner, 73, 205 INDEX OF SUBJECTS. A PAGES A-line.. ii, 18, 66 Abnormal dispersion 15 Absorption and emission of light, interrelationship 6, 76, 197 fluorescence 79 phosphorescence . . 79 bands, displacement produced by increasing atomic weight 132, 181 substitution in organic com- pounds 180,181 coefficient 79 of light by gases and liquids 175 in the infra-red . 184 ultra-violet 182, 202 visible region of the spectrum 179 mechanism .... 77 -spectra i, 63, 76, 174 and color, interrelationship 180 molecular structure, interrelationship 78, 178 influence of concentration of a solution 177 increasing atomic weight 132, 181 optical density 79 position of substituting groups 180 solvent 78 state of aggregation 77 substituting radicles 179, 182 temperature 77 observation of 63, 1 74 photometric 39 table of : 176 thickness of layer of active substance 64, 79 Aggregation, influence on absorption-spectrum 78 Air, absorption-spectrum ........ 151 221 222 INDEX OF SUBJECTS. Air, line-spectrum 149, ii>i lines in metallic spectra 57, 149 liquid, spectrum of electric discharge in 1 50 refractive index 13, 67 Albumins, ultra-violet absorption-spectra 184 Albuminoids, ultra-violet absorption-spectra 184 Alcohol, refractive index 13 Alcohols C M H 2 +i.OH ultra-violet absorption-spectra 182, 183 Aldebaran spectrum 209 Alizarin absorption-spectrum 1 76 Alumina arc (band) spectrum 97 Aluminium 96 arc-spectrum 96 spark-spectrum 96 Amines 15 Ammonia infra-red absorption-spectrum 184 line (flame) spectrum 149, 152 Angle, refractive 13 Angstrom's unit 12 Aniline-blue absorption-spectrum 176 Antimony 97 absorption-spectrum. 98 arc-spectrum 98 spark-spectrum 98 Arc-spectra, production of 54 Arcturus spectrum 209 Argand burner 63 Argon 98 blue (spark) spectrum 100 red (spark) spectrum 100 white (spark) spectrum 101 Arsenic 101 arc-spectrum 102 spark-spectrum 102 Astigmatism of concave grating 48 Atlas, Angstrom's 6, 188, 189 Cornu's 7, 188- Rowland's 7, 189 Atmosphere, terrestrial, absorptive action 151, 1 86, 201, 202 Atomic and luminiferous vibrations, interrelationship 70 vibrations of molecules 70, 73 weight and spectra, interrelationship 87, 88, 91 calculation of, from homologous lines 88 Aurora borealis spectrum 211 Azo-group, influence on organic dyes 180, 1.81 INDEX OF SUBJECTS 22$ B PAGES -\me 1 8, 66 /Mine 65, 206, 207 ^2-, ^s-, ^4- line 66, 206 Band-spectra 69 formula for calculation of (Deslandres) Si, 86 regularities in construction of 86 Barium 102 arc-spectrum 1 03 spark-spectrum 103 Barium bromide flame-spectrum 103 chloride flame-spectrum 103 iodide flame-spectrum 103 oxide flame-spectrum . 103 Basic lines 198 Benzene and derivatives ultra-violet absorption-spectra 182 Beryllium 103 arc-spectrum 103 fluorescent spectrum 103 spark-spectrum 103 Bessemer converter flame-spectrum 1 36 Bismuth 104 arc-spectrum 104 spark-spectrum 104 oxide flame-spectrum 104 salts flame-spectrum , . . . . 104 Blood-pigments absorption-spectra > 176 stains, identification 175 Bolometer (Langley's) 62, 197 Boric acid flame (band) spectrum 105 Boron 104 arc-spectrum 105 spark-spectrum 105 Broadening of spectrum lines 73, 75 Bromine 105 absorption-spectrum 106 spark (line) spectrum 106 a-Bromonaphthalene coefficient of refraction , . . . 13 use 44 Burner, Auer's 44, 63 Bunsen's 51, 54 Terquem's 51, 53 224 INDEX OF SUBJECTS. C PAGFS C-line 18, 66, 206 Cadmium 107 arc-spectrum 108 spark-spectrum 108 bromide flame-spectrum 108 chloride flame-spectrum 108 Caesium 108 arc-spectrum 108 flame-spectrum 108 spark-spectrum 1 08 Calcium 109 arc-spectrum 109 flame-spectrum 109 spark-spectrum 109 bromide flame-spectrum no chloride flame-spectrum no fluoride flame-spectrum no oxide flame-spectrum no Calculation of the position of spectrum-lines (Kayser and Runge) 82 (Rydberg) 85 Capella spectrum . 209 Carbohydrates ultra-violet absorption-spectrum 184 Carbon no band (flame, arc) spectrum in, 113 line (spark) spectrum 113 bands in metallic spectra 112 presence in comets 210 stars 209 bisulphide refractive index 13 use in hollow prisms 27 monoxide band (spark) spectrum 114 toxicological detection 1 76 haemoglobin 176 X-Cassiopae spectrum 208 Cathochromic groups , 1 8 1 Cerium 116, 201 spark-spectrum 116 Channelled spectra 69, 71 Chlorine , 116 absorption-spectrum 117 spark-spectrum 117 Chlorophyll absorption-spectrum 1 76 Chromium 117 UNIVERSITY INDEX OF SUBJECTS. 225 PAUES ' Chromium arc-spectrum 118- spark-spectrurn 117 compounds absorption-spectrum 118, 176- spectro-analytical determination 1 18 Chromophors , 181 Chromosphere 206 Chrysoldins absorption-spectra 179 Classification of colors (Listing) 95, Cinnamate ethylic refraction coefficient 13 use in hollow prisms 28 Cobalt 1 18 arc-spectrum 119 spark-spectrum 119 chloride and ammonium thiocyanate 119 glass absorption-spectrum 119 salts absorption-spectrum 119, 176 spectro-analytical determination 119 Coefficient of extinction 79 Coincidence method (Rowland) 49 Collimator 22, 48 tube 21 Color and absorption-spectrum, interrelationship 1 80 of salt solutions, conformation with Arrhenius' dissociation hy- pothesis 177 Colors, classification of (Listing) 95 complementary 180 combined produce whiteness 14, 180- explanation 1 80 wave-lengths 95 Comets spectra 210 self-luminosity 210 Comparison photographs 48 spectrum 22 Complementary colors 180- Concave grating (Rowland) 7, 34, 45, 47. 63, 189 spectrograph (Rowland) 45 spectroscope (Rowland) 33 Continuous spectra i, 69 ' Copper 121 arc-spectrum .^ 122 spark-spectrum 122 chloride flame-spectrum 122 salts absorption-spectrum 121, 176 spectro-analytical determination 121 Cooling of flames for production of spectra 51 226 INDEX OF SUBJECTS. PAGES Corona 207 Coronal lines 202, 206, 207 7-Coronae spectrum ... 209 Correction table (Watts) for Angstrom's scale 94 Crown-glass refraction coefficient 13 Cyanogen band-spectrum 112, 114 D ZMines 4, n, 16, 18, 76, 207 ZVline wave-length n, 18, 49, 66, 190, 206 ZVline wave-length 66, 206 Z> 3 -line 126,202, 206 Delicacy of spectrum reactions . . '. 95 Demonium spectrum 201 Deviation, minimum 14 Diagrams of spectra 68 Didymium. 122 spark-spectrum 123 chloride absorption-spectrum 123, 124 earths absorption-spectrum 123 nitrate absorption-spectrum , 1 24 Diffraction grating, see Grating. discovery by Fraunhofer 17 of light 17 Discontinuous spectra I Dispersion, abnormal 15 curve 66 formula (Cauchy) 37 increase ol 23 of light 14 Displacement of lines 212 Dissociation, electrolytic 79, 132 hypothetical, of elements in the sun. . . ., 198 Doppler's principle , 212 Double lines 82 Dyes, organic, properties dependent on the presence of specific groups in the molecule 180 theory of (Schiitze) i&o E ^-line.... 18.66 Elements present in the sun 197 Emission and absorption of light, interrelationship 76 of light i spectra 69 INDEX OF SUBJECTS. 22$ PAGES Eosin absorption-spectrum 176, 180 Erbium 124, 201 spark-spectrum 124 chloride absorption-spectrum 125 Esters, see Ethereal salts. Ether, luminiferous.. n ultra-violet absorption-spectra 183 Ethereal salts, ultra-violet absorption-spectra 182 Eihylic cinnamate refractive index 13 Extinction coefficient , 79 Eyepiece (Gauss) 37 F T^-line ...'. 66, 206, 212 Faculae 203, 205 Fatty acids ultra-violet absorption-spectrum 182 Fixed stars spectrum 208 Flame-spectra, apparatus for producing 52 Flint-glass refraction coefficient. . . 13 Fluorescein absorption-spectrum. 176 Fluorescence and absorption 79 Fluorine 125 flame-spectrum 125 spark-spectrum 125 Fluor-spar refraction-coefficient 13 Formula, Balmer's 81, 82 Deslandres' , .. 81, 86 Fraunhofer lines 76, 151, 186, 189, 191, 200 Fuchsin (magenta) absorption-spectrum 176 G <7-line *..... 18, 66, 207 Gallium . . 125 arc-spectrum . . 125 spark-spectrum 125 spectro-analytical calculation of atomic weight 89 Gases, spectroscopic investigation of 58 Gauss' eyepiece 37 Geissler tubes , , ^ 58 filling of 58 Germanium. . . . 125 arc-spectrum. ...... , 126 spark-spectrum 126 spectro-analytical calculation of atomic weight. 88 228 INDEX OF SUBJECTS. PAGES Glass, contamination of spectra by 26 Schultz's 39 gratings 17 Gold 126 arc-spectrum 126 spark-spectrum 126 Graphic representation of spectra 68 Grating 16 constant 32 holder 47 plane and concave (Rowland) 17, 33, 63 reflection 17 separatory power 19 spectra, 'production of 31 spectroscope 31, 45. transparent 17 H J/-line 18, 66, 206 Haematin absorption-spectrum 176 Half-prisms (Christie) 31 Harmonic relationship of spectrum lines 8, 80 series (Liveingand Dewar) 8, 81 Heating effect of various regions of the solar spectrum 61 Helium 126 spark-spectrum 127 ft- Hercules spectrum 209 H istory of spectrum analysis 2 Hollow prisms for the observation of absorption-spectra 64 Homologous lines of different elements 87 Hydrogen I2& compound line-spectrum 130 elementary line-spectrum 130 infra-red lineal absorption-spectrum 184 flame 51 lines, formula for calculating (Balmer) 81, 82, 129 Hypsochromic groups 181 I Incandescent burner ( Auer's) 44, 63 Indigo absorption-spectrum 1 76 Indium , 131 arc-spectrum 131 flame-spectrum 131 spark-spectrum 131 INDEX OF SUBJECTS. 22$ PACKS Induction-spark 55 Influence machine 55 Infra-red (ultra-red) rays 61 method of observation 61 designation of lines in 189 Instruments for spectrum analysis 20 Intensity numbers 95 scales. ... 95 of spectrum-lines 95 Interference, law of 18 Interpolation-curve construction 66 Iodine. 132 absorption-spectrum . . r 133 spark-spectrum 133 compounds absorption-spectrum 133 Ions 177 Iridium 134 arc-spectrum 1 34 ammonium chloride absorption-spectrum 134 Iron 135 arc-spectrum 137 spark-spectrum 135 salts absorption-spectrum 135, 176 Isomeric compounds absorption-spectra 183 J Jena glass refractive index 13 Jupiter spectrum 209 Jupiter's satellites spectrum , 209 K AMine ' n, 66, 206 Kirchhoff's law 6, 197 Kundt's rule 78, 178 L Lamp (Bartel's) 51 Lanthanum 141, 201 spark-spectrum 141 Law of exchanges (Kirchhoff) 6, 76, 197 objections to , 77 Layer, reversing. . . 198, 203, 205 Lead 141 arc-spectrum 142 230 INDEX OF SUBJECTS. PAGES Lead spark-spectrum 142 oxide band-spectrum . ; 141 Leyden jar 55 Light, anomalous dispersion 15 diffraction 17 dispersion 14 homogeneous : 15 reflection 12 refraction 12 theory of (Huygens) n velocity n, 212 vibrations, relation to atomic vibrations 70 white, composition (Newton) 15 Lightning spectrum 211 Line 1474 K (coronal line) 202. 206, 207 displacement 212 of collimation, motion of luminous bodies in 212 pairs 82, 91 spectra , . . . . 69 triplets 85, 91 Lines, basic (suppositive) i q8 Fraunhofer 76, 151, 186, 189, 191, 200 origin 6, 77 homologous 87 long and short (Lockyer) 73, 198 .of a spectrum, formula for the calculation of (Kayser and Runge). . . 82 (Rydberg) 84 an element, interrelationship 80 spontaneous reversal 72 widening by increase of pressure 72 in magnetic current. 75 Lithium 142 arc-spectrum , j 42 flame-spectrum 142 spark-spectrum 142 Luminescence (Wiedemann) 77 Luminiferous ether n Luminous paint (Balmain's) 63 a-Lyrae spectrum 208 M Magenta (fuchsin) absorption-spectrum 176 Magnesium 142 arc-spectrum 143. INDEX OF SUBJECTS. PAGES Magnesium flame-spectrum 143, spark-spectrum 143, hydride band-spectrum 144: oxide band-spectrum 144. Magnetic current, influence on spectrum-lines 76 Malachite green absorption-spectrum 176 Manganese 144 arc-spectrum 145, spark-spectrum 145; oxide spark-spectrum 145 salts absorption-spectrum 144, 176 spectro-colorimetric determination 144^ Mars spectrum 2091 Measuring appliances, spectroscopic 65, 67 Mercury 145. arc-spectrum 145 spark-spectrum 145,. (planet) spectrum 2og Metal gratings 31 Meteors and shooting stars , 210 Methyl violet absorption-spectrum 176 Micromillimetre (/n/j) 12 Microspectroscope 40 Miniature spectroscope 29 Minimum of brightness, determination by Kriiss* method 64, 174. deviation *. . 14 automatic adjustment in position of 24 Molecular structure and absorption-spectrum relationship 178, 179, 182 of matter, investigation of 69 Molecules, atomic movement of . 70 of gases 69, 70 liquids 69, 70 solids 69, 70 Molybdenum 146 spark-spectrum 146 Monckhoven's tubes 59 a-Monobromonaphthalene in hollow prisms 44 Moon spectrum 209 Motion of luminous bodies in the line of collimation , 213, N Nebulae spectrum 210* Nebulous series (Rydberg) 285 Neodymium 1 22. 232 INDEX OF SUBJECTS. I'AGF.S Neptune spectrum 209 Nickel 146 arc-spectrum 147 spark-spectrum 147 salts absorption-spectrum 146, 176 Niobium 147 Nitric acid infra-red absorption-spectrum 184 Nitrogen 148 band-spectrum (-f- pole) 151 (- pole) 151 line-spectrum 1 50 spectrum of electric discharge in liquid 1 50 Nitrous anhydride (NO -f- NO 2 ) absorption-spectrum 150, 152 Normal lines, Rowland's table of 191 O Orientation of spectra 113, 135 a-Orionis spectrum 209 /S-, Y't $-, 6-Orionis spectra 208 Oscillation frequency 67 Osmium 153 arc-spectrum 153 spark-spectrum 153 Oxygen 153 band-spectrum , . 154 compound line-spectrum 154 elementary line-spectrum 154 inorganic radicles infra-red absorption-spectrum 1 84 liquid, spectrum of electric discharge in 150 Oxyhaemoglobin absorption-spectrum 1 76 Oxyhydrogen flame , 51 P Pairs 82, 91 Palladium 1 54 arc-spectrum 155 spark-spectrum 155 /f-Pegasi spectrum 209 Phosphorus 155 band (flame) spectrum 155 line (spark) spectrum 155 Phosphorescence 79 Phosphorescent action of infra-red rays 63 Photographic plates (Abney) 62, 188 (Schumann) 61 INDEX OF SUBfECTS. 233 PACKS Photographic plates without gelatine for ultra-violet rays 6r Photography, spectroscopic use of 44, 47, 62, 180 Photosphere 203, 204 Planets spectrum 209 Platinum 156 arc-spectrum 156 spark-spectrum 156 salts absorption-spectrum 156 Plucker's tubes 58 Pocket spectroscope 29 Potassium 1 56 arc-spectrum 157 flame-spectrum 1 56 spark-spectrum 157 Potsdam system 7 Praseodymium spectrum 201 nitrate absorption-spectrum 124 Pressure, influence on spectrum lines 72 Principal series (Kayser and Runge) 83 (Rydberg) 85 Prism material, coefficients of refraction t 13 spectroscopes with angular vision 20 direct vision 28 Prisms 13 compound (Amici) 28 hollow 24 refracting angle 14 edge 14 faces 13 Rutherfurd's 26, 44 separatory power 16 totally reflecting 25 Prominences, solar 42, 205 eruptive 206 quiescent 205 velocity 206, 213 s-Puppis spectrum t 129 Purpurin absorption-spectrum 176 Quantitative spectrum analysis 39, 118, 119, lai, 144 Quartz lenses and prisms 61 refractive coefficient , 13 Quinine sulphate for observation of ultra-violet rays 60 Qu incline red absorption -spectrum 176 234 INDEX OF SUBJECTS. R PAGES Rain-band 129 Rare earths spectra 201 ultra-violet absorption-spectra 122 Rays, infra-red - 61 ultra-violet 60 Red sta rs 209 Reduction of Angstrom's to Rowland's scale 94 scale-measurements to wave-lengths 66 Reflection of light 12 prism 22 Refraction coefficient 13 index 13 of light. 12 Regularities in the construction of spectra 8, 80 Relationship between the lines of an element 8, 80 spectra of different elements 8,87 Representation of spectra 68 Reversal of spectrum lines 72, 203 Reversing layer of solar atmosphere 198, 203, 205 Rhodium 157 arc-spectrum 157 Rubidium , 158 arc-spectrum 15$ flame-spectrum 15& spark-spectrum 158 Rule, Kundt's 78, 178 Ruthenium 158 arc- spectrum , 158 S Safranine absorption-spectrum 176 Salet's tubes 59 Salt solutions, color of, and Arrhenius' dissociation theory 177 Salts with a common colored basic or acidic radicle 177 Samarium ! 58 spark-spectrum 159. nitrate absorption-spectrum 159 Saturn spectrum 209 Srale, Angstrom's u ) 92 Bunsen's. 65, 66 Rowland's 66, 92 INDEX OF SUBJECTS. 235 PAGES Scale divisions, conversion into wave-lengths 66 Scales 65, 67, 92 inequality of 65 Scandium 159 spark-spectrum 159 Schultz's glasses ' 39 Selenium. ... , 159 band-spectrum 160 spark-spectrum 160 compounds absorption-spectrum 160 Separatory power of a grating 19 prism 19 Series, harmonic 81 of lines, principal, nebulous, sharp (Rydberg) 85 of an element 81, 84 Silicious flint-glass refractive index 13 Silicon 160 spark-spectrum 160 fluoride spark (flame) spectrum 125 Silver 161 arc-spectrum 161 spark-spectrum 161 Sirius motion in line of collimation 213 spectrum 208 Slit construction 22, 48 introduction 16 tube (collimator) 22, 48 wedge-shaped 44 Snell's law 13 Sodium 161 arc-spectrum 162 flame-spectrum 161 spark-spectrum 162 (D) line 4, n, 16, 18, 76, 207 Solar atlas, Angstrom's 6, 188, 189 Cornu's (ultra-violet) 7, 188 Rowland's 7, 189 atmosphere, presence of terrestrial substances in 197 velocity 206, 2 1 2 faculae 203, 205 nucleus 203, 204 prominences 42, 205 spectroscopes 41 spectrum 6 absorption of ultra-violet region 202 236 INDEX OF SUBJECTS. PAGES Solar spectrum Cornu's atlas of ultra-violet region 7, 188 Kirchhoff's drawings 188 normal, Angstrom's atlas 6, 188, 189 Rowland's photographic atlas 7, 189 photography of infra-red region 62, 1 88 telluric lines. ... 151, 186, 201 Solutions, spectroscopic examination 64 Spark gap, introduction into circuit 55 spectra 3 production 55 Spectra, diffraction, merits of 19 grouping in agreement with periodic system 82 influence of magnetic current 75 pressure 72 temperature , 72 of different elements, interrelationship. 8, 87, 89 electric spark, first observation 3 first order (Pliicker and Hittorf) 71 second order Pliicker and Hittorf) 71 recording of (Bunsen) 68 in accordance with number of vibrations 90 refraction, merits .. 19 relationship to atomic weights 83, 91 Spectrographs 43 Spectrometers 35 Spectrophotometer 39 Spectroscopes angular vision 20 direct vision 28 grating 31 Spectroscopic charts 188 instruments 35 Spectrum i, 92 absorption ... 174 Spectrum analysis, applications I, 8, 40, 92, 1 36 foundation by Bunsen and Kirchhoff 2, 5 general bibliography S history 2 physical basis 1 1 province i quantitative 39, 118, 119, 121, 144 band 69 change of, dependent on alteration in atomic vibration 70 channelled space 69. 71 constancy in the same order : 71 continuous i , 69 INDEX OF SUBJECTS. 2$? PAGBS Spectrum definition I discontinuous i invisible, observation Go line 69 lines, widening 72, 75, 205 of an element different from that of its compounds 5, 71 electric discharge in highly rarefied gases (Plucker) 5 orientation 6, 67, 135, 186 pure 16 reactions, delicacy 95 Spontaneous reversal of spectrum-lines 72 Stars spectrum 208 red 209 shooting and meteors 210 white 208 Stellar spectrometers 43 spectroscopes 41 Strontium 162 arc-spectrum 163 spark-spectrum 163 chloride flame-spectrum 163 compounds flame-spectrum 162 oxide flame-spectrum 163 Sub-series, first and second (Kayser and Runge) 82 Substituting groups in organic compounds 171; influence on position of absorption-bands 179 Sulphur. 163 band-spectrum 164 line-spectrum 164 Sun chemical composition 190 light as an illuminant 63 physical condition v 203 spots 203, 204 Survey, spectroscopic, of the sky 43 Swan's carbon-spectrum no Symmetrical path of rays through a prism 14 T Fable for reduction of wave-lengths to Rowland's scale 94 Tantalum 164. Telluric lines in solar spectrum 151, 186, 201 Tellurium 1 64 band-spectrum 165 spark-spectrum 1 03 238 INDEX OF SUBJECTS. Tellurium bromide absorption-spectrum 165 chloride absorption-spectrum 165 Temperature influence on absorption-spectrum 77 spectrum lines 72 of Bunsen flame 54 electric arc 54 spark 55 Terrestrial lines of solar spectrum, see Telluric lines. Thallium 165 arc-spectrum 166 flame-spectrum 1 66 spark-spectrum 166 Thorium 166, 201 spark-spectrum 166 Thulium 167 spark-spectrum 167 oxide band-spectrum 167 salts absorption-spectrum 167 Tin 167 arc-spectrum 168 spark-spectrum 168 oxide band-spectrum. 167 Titanium 168 arc-spectrum 168 spark-spectrum 168 Triplets 85, 89, 91 Tubes, Geissler's 58 Monckhoven's 59 Salet's 59 Tungsten 169 spark-spectrum 1 70 U Ultra-red rays, see Infra-red rays. Ultra-violet rays 60 method of observation 6 O Unit, Angstrom's 12 Universal spectroscope (Kriiss) 38 Uranium 170 spark-spectrum 1 70 glass, use for observation of ultra-violet rays 60 Uranium salts absorption-spectrum , .. , . 170, 176 Uranus spectrum 209 INDEX OF SUBJECTS. 239 V PAGES Vacuum-tubes, filling 'of 58 Vanadium 171 arc-spectrum 171 spark-spectrum 171 Vapor, production of glowing , 51 Vega spectrum 208 Velocity of light 213 Venus spectrum 209 W Water absorption-spectrum , 130 in infra-red 184 refractive index 13 vapor absorption-spectrum 129 Wave-length u, 188 change in, by the approach or recession of the source of light. 212 determinations of the /Mines 92 of the colors 95 table of the Fraunhofer lines (Rowland). 191 unit II, 92 use in orientation of spectra 190 White stars 208 Widening of spectrum lines 72, 75, 205 Y Ytterbium 171 spark-spectrum 171 Yttrium 171, 201 phosphorescent spectrum 171, 172 spark-spectrum 172 Z Zinc 172 arc-spectrum 172 spark-spectrum 172 Zirconium.. . . 172 light (Linnemann's) 63 spark-spectrum... . , 173 Zodiacal-light spectrum 211 SHORT-TITLE CATALOGUE OF THE PUBLICATIONS OF JOHN WILEY & SONS, NEW YORK, LONDON: CHAPMAN & HALL, LIMITED. ARRANGED UNDER SUBJECTS. Descriptive circulars sent on application. Books marked with an asterisk are sold at net prices only. All books are bound in cloth unless otherwise stated. AGRICULTURE. CATTLE FEEDING DAIRY PRACTICE DISEASES OF ANIMALS GARDENING, ETC. Armsby's Manual of Cattle Feeding, 12mo, $1 75 Downing's Fruit and Fruit Trees Svo, 5' 00 Grotenfelt's The Principles of Modern Dairy Practice. (Woll.) 12mo, 2 00 Kemp's Landscape Gardening 12mo, 2 50 Lloyd's Science of Agriculture Svo, 4 00 Loudon's Gardening for Ladies. (Downing.) 12mo, 1 50 Steel's Treatise on the Diseases of the Dog Svo, 3 50 " Treatise on the Diseases of the Ox Svo, 600 Stockbridge's Rocks and Soils Svo, 2 50 Woll's Handbook for Farmers and Dairymen 12mo, 1 50 ARCHITECTURE. BUILDING CARPENTRY STAIRS VENTILATION, ETC. Berg's Buildings and Structures of American Railroads 4to, 7 50 Birkmire's American Theatres Planning and Construction. Svo, 3 00 Architectural Iron and Steel Svo, 3 50 f ",<> Compound 'Riveted Girders Svo, 2 00 Skeleton Construction in Buildings Svo, 3 00 Planning and Construction of High Office Buildings. 1 Carpenter's Heating and Ventilating of Buildings 8vo, $3 00 Downing, Cottages Svo, 2 50 " Hints to Architects ,Svo, 2 00 Freitag's Architectural Engineering Svo, 2 50 Gerhard's Sanitary House Inspection 16mo, 1 00 Theatre Fires and Panics 12mo, 1 50 Hatfield's American House Carpenter Svo, 5 00 Holly's Carpenter and Joiner ISmo, 75 Kidder's Architect and Builder's Pocket-book Morocco flap, 4 00 Merrill's Stones for Building and Decoration Svo, 5 00 Monckton's Stair Building Wood, Iron, and Stone 4to, 4 00 Wait's Engineering and Architectural Jurisprudence Svo, 6 00 Sheep, 6 50 Worcester's Small Hospitals Establishment and Maintenance, including Atkinson's Suggestions for Hospital Archi- tecture 12mo, 125 World's Columbian Exposition of 1893 4to, 2 50 ARMY, NAVY, Etc. MILITARY ENGINEERING ORDNANCE PORT CHARGES, ETC. Bourne's Screw Propellers .4to, 5 00 Bruff s Ordnance and Gunnery Svo, 6 00 Buckuill's Submarine Mines and Torpedoes Svo, 4 00 Chase's Screw Propellers Svo, 3 00 Cooke's Naval Ordnance Svo, 12 50 Cronkhite's Gunnery for Non-coin. Officers ISmo, morocco, 2 00 De Brack's Cavalry Outpost Duties. (Carr.). . . .ISmo, morocco, 2 00 Dietz's Soldier's First Aid 12mo, morocco, 1 25 * Dredge's Modern French Artillefy 4to, half morocco, 20 00 Record of the Transportation Exhibits Building, World's Columbian Exposition of 1893.. 4to, half morocco, 10 00 Dyer's Light Artillery 12mo, 3 00 Hoff's Naval Tactics Svo, 1 50 Hunter's Port Charges Svo, half morocco, 13 00 Ingalls's Ballistic Tables Svo, 1 50 Handbook of Problems in Direct Fire Svo, 4 00 Mahau's Advanced Guard ISmo, 1 50 " Permanent Fortifications. (Mercur.).8vo, half morocco, 7 50 Mercur's Attack of Fortified Places 12mo, 2 00 2 Mercur's Elements of the Art of War 8vo, $4 00 Metcalfe's Ordnance and Gunnery 12uio, with Atlas, 5 00 Murray's A Manual for Courts-Martial 18mo, morocco, 1 50 " Infantry Drill Regulations adapted to the Springfield Rifle, Caliber .45 18mo, paper, 10 Phelps's Practical Marine Surveying 8vo, 2 50 Powell's Army Officer's Examiner 12mo, 4 00 Reed's Signal Service 50 Sharpe's Subsisting Armies 18mo, morocco, 1 50 Todcl and Whall's Practical Seamanship 8vo, 7 50 Very's Navies of the World 8vo, half morocco, 3 50 Wheeler's Siege Operations 8vo, 2 00 Wiuthrop's Abridgment of Military Law 12mo, 2 50 Woodhull's Notes on Military Hygiene 12mo, morocco, 2 50 Young's Simple Elements of Navigation.. 12mo, morocco flaps, 2 50 ASSAYING. SMELTING ORE DRESSING ALLOYS, ETC. Fletcher's Quant. Assaying with the Blowpipe.. 12mo, morocco, 1 50 Furnian's Practical Assaying 8vo, 3 00 Kuuhardt's Ore Dressing 8vo, 1 50 * Mitchell's Practical Assaying. (Crookes.) 8vo, 10 00 O'Driscoll's Treatment of Gold Ores 8vo, 2 CO Ricketts and Miller's Notes on Assaying 8vo, 3 00 Thurston's Alloys, Brasses, and Bronzes, . , 8vo, 2 50 Wilson's Cyanide Processes l"2rno, 1 50 The Chlorination Process 12mo, 150 ASTRONOMY. PRACTICAL, THEORETICAL, AND DESCRIPTIVE. Craig's Azimuth 4to, 3 50 Doolittle's Practical Astronomy 8vo, 4 00 Gore's Elements of Geodesy 8vo, 2 50 Michie and Harlow's Practical Astronomy 8vo, 3 00 White's Theoretical and Descriptive Astronomy 12mo, 2 00 BOTANY. GARDENING FOR LADIES, ETC. Baldwin's Orchids of New England 8vo, 1 50 Loudon's Gardening for Ladies. (Downing.) 12mo, 1 50 3 MacCord's Kinematics ... 8vo, $5 00 " Mechanical Drawing 8vo, 400 Mahan's Industrial Drawing. (Thompson.) 2 vols., 8vo, 3 50 Reed's Topographical Drawing. (II. A.) 4to, 5 00 Smith's Topographical Drawing. (Macmillan.) 8vo, 2 50 Warren's Descriptive Geometry 2 vols., 8vo, 3 50 " Drafting Instruments . . '. ...12mo, 1 25 ' ' Free-hand Drawing 12mo, 1 00 " Higher Linear Perspective 8vo, 3 50 " Linear Perspective 12mo, 100 " Machine Construction. 2 vols., 8vo, 7 50 " Plane Problems , 12mo, 125 " Primary Geometry . ...12mo, 75 " Problems and Theorems 8vo, 2 50 " Projection Drawing . . . 12mo, 1 50 " Shades and Shadows 8vo, 3 00 " Stereotomy Stone Cutting 8vo, 2 50 Whelpley's Letter Engraving ... 12mo, 2 00 ELECTRICITY AND MAGNETISM. ILLUMINATION BATTERIES PHYSICS. Anthony and Brackett's Text-book of Physics (Magie). ...8vo, 400 Barker's Deep-sea Soundings. 8vo, 2 00 Benjamin's Voltaic Cell 8vo, 3 00 History of Electricity 8vo 3 00 Cosmic Law of Thermal Repulsion ;..... 18mo, 75 Crehore and Squier's Experiments with a New Polarizing Photo- Chronograph 8vo, 3 00 * Dredge's Electric Illuminations. . . .2 vols. , 4to, half morocco, 25 00 Vol. II...' 4to, 7 50 Gilbert's De magnete. (Mottelay.) 8vo, 2 50 Holman's Precision of Measurements 8vo, 2 00 Michie's Wave Motion Relating to Sound and Light Svo, 4 00 Morgan's The Theory of Solutions and its Results 12mo, 1 00 Niaudet's Electric Batteries. (Fishback.) . . 12mo, 2 50 Reagan's Steam and Electrical Locomotives 12mo 2 00 Thtirston's Stationary Steam Engines for Electric Lighting Pur- poses 12mo, 1 50 Tillman's Heat 8vo, 1 50 ENGINEERING. CIVIL MECHANICAL SANITAKY, ETC. (See also BRIDGES, p. 4 ; HYDRAULICS, p. 8 ; MATERIALS OF EN- GINEERING, p. 9 ; MECHANICS AND MACHINERY, p. 11 ; STEAM ENGINES AND BOILERS, p. 14.) Baker's Masonry Construction 8vo, $5 00 Baker's Surveying Instruments 12mo, 3 00 Black's U. S. Public Works 4to, 5 00 Brook's Street Railway Location , 12mo, morocco, 1 50 Butts's Engineer's Field-book 12mo, morocco, 2 50 Byrne's Highway Construction 8vo, 7 50 Carpenter's Experimental Engineering. 8vo, 6 00 Church's Mechanics of Engineering Solids and Fluids 8vo, 6 00 " Notes and Examples in Mechanics 8vo, 2 00 Crandall's Earthwork Tables 8vo, 1 50 Crandall's The Transition Curve 12mo, morocco, 1 50 * Dredge's Penn. Railroad Construction, etc. . . Folio, half mor., 20 00 * Drinker's Tunnelling 4to, half morocco, 25 00 Eissler's Explosives Nitroglycerine and Dynamite 8vo, 4 00 Gerhard's Sanitary House Inspection .16mo, 1 00 Godwin's Railroad Engineer's Field-book. 12mo, pocket-bk. form, 2 50 Gore's Elements of Geodesy 8vo, 2 50 Howard's Transition Curve Field-book 12mo, morocco flap, 1 50 Howe's Retaining Walls (New Edition.) , 12mo, 1 25 Hudson's Excavation Tables. Vol. II 8vo, 1 00 Button's Mechanical Engineering of Power Plants 8vo, 5 00 Johnson's Materials of Construction 8vo, 6 00 Johnson's Stadia Reduction Diagram. .Sheet, 22 X 28 inches, 50 Theory and Practice of Surveying 8vo, 4 00 Kent's Mechanical Engineer's Pocket-book 12mo, morocco, 5 00 Kiersted's Sewage Disposal 12mo, 1 25 Kirkwood's Lead Pipe for Service Pipe 8vo, 1 50 Mahan's Civil Engineering. (Wood.) 8vo, 5 00 Merriman and Brook's Handbook for Surveyors. . . .12mo, mor., 2 00 Merriman's Geodetic Surveying 8vo, 2 00 " Retaining Walls and Masonry Dams 8vo, 2 00 Mosely's Mechanical Engineering. (Mahan.) 8vo, 5 00 Nagle's Manual for Railroad Engineers 12mo, morocco, 3 00 Pattern's Civil Engineering 8vo, $7 50 " Foundations 8vo, 500 Rockwell's Roads and Pavements in France 12mo, 1 25 Ruffiier's Non-tidal Rivers 8vo, 1 25 Searles's Field Engineering 12mo, morocco flaps, 3 00 " Railroad Spiral 12mo, morocco flaps; 1 50 Siebert and Biggin's Modern Stone Cutting and Masonry. . .8vo, 1 50 Smith's Cable Tramways 4to, 2 50 " Wire Manufacture and Uses 4to, 3 00 Spaldiug's Roads and Pavements 12mo, 2 00 " Hydraulic Cement 12mo, 2 00 Thurstou's Materials of Construction, 8vo, 5 00 * Trautwiue's Civil Engineer's Pocket-book. ..12mo, uior. flaps, 5 00 * " Cross-section Sheet, 25 * " Excavations and Embankments 8vo, 2 00 * " Laying Out Curves 12mo, morocco, 2 50 Waddell's De Pontibus (A Pocket-book for Bridge Engineers.) Wait's Engineering and Architectural Jurisprudence 8vo, 6 00 Sheep, 6 50 Warren's Stereotomy Stone Cutting 8vo, 2 50 Webb's Engineering Instruments 12mo, morocco, 1 00 Wegmann's Construction of Masonry Dams 4to, 5 00 Wellington's Location of Railways 8vo, 5 00 Wheeler's Civil Engineering 8vo, 4 00 Wolff's Windmill as a Prime Mover 8vo, 3 00 HYDRAULICS. WATER-WHEELS WINDMILLS SERVICE PIPE DRAINAGE, ETC. (See also ENGINEERING, p. 6.) Bazin's Experiments upon the Contraction of the Liquid Vein (Trautwine) 8vo, 2 00 Bovey's Treatise on Hydraulics 8vo, 4 00 Coffin's Graphical Solution of Hydraulic Problems 12mo, 2 50 Ferrel's Treatise on the Winds, Cyclones, and Tornadoes. . .8vo, 4 00 Fuerte's Water and Public Health 12mo, 1 50 Ganguillet & Kutter's Flow of Water. (Heriug&Trautwiue.).8vo, 4 00 Hazeu's Filtration of Public Water Supply 8vo, 2 00 Herschel's 115 Experiments 8vo, 2 00 Kiersted's Sewage Disposal 12mo, 1 25 Kirk wood's Lead Pipe for Service Pipe 8vo, $ 1 50 Masou's Water Supply 8vo, 5 00 Merrimaii's Treatise oil Hydraulics. . - , 8vo, 4 00 Nichols's Water Supply (Chemical and Sauitary) 8vo, 2 50 Ruffner's Improvement for Non-tidal Rivers 8vo, 1 25 Wegmaun's Water Supply of the City of New York 4to, 10 00 Weisbach's Hydraulics. (Du Bois.) 8vo, 5 00 Wilsou's Irrigation Engineering ... .8vo, 4 00 Wolff's Windmill as a Prime Mover 8vo, 3 00 Wood's Theory of Turbines 8vo, 2 50 MANUFACTURES. ANILINE BOILERS EXPLOSIVES IRON SUGAR WATCHES WOOLLENS, ETC. Allen's Tables for Iron Analysis 8vo, 3 00 Beaumont's Woollen and Worsted Manufacture 12ino, 1 50 Bolland's Encyclopaedia of Founding Terms 12mo, 3 00 The Iron Founder 12mo, 250 " " Supplement 12mo, 250 Booth's Clock and Watch Maker's Manual 12mo, 2 00 Bouvier's Handbook on Oil Painting .12mo, 2 00 Eissler's Explosives, Nitroglycerine and Dynamite 8vo, 4 00 Ford's Boiler Making for Boiler Makers 18mo, 1 00 Metcalfe's Cost of Manufactures 8vo, 5 00 Metcalf 's Steel A Manual for Steel Users 12mo, 2 00 Reimaun's Aniline Colors. (Crookes.) 8vo, 2 50 * Reisig's Guide to Piece Dyeing 8vo, 25 00 Spencer's Sugar Manufacturer's Handbook. . . .12mo, inor. flap, 2 00 " Handbook for Chemists of Beet Houses. 12mo, inor. flap, 3 00 Svedelius's Handbook for Charcoal Burners 12mo, 1 50 The Lathe and Its Uses 8vo, 6 00 Thurston's Manual of Steam Boilers 8vo, 5 00 Walke's Lectures on Explosives 8vo, 4 00 West's American Foundry Practice 12mo, 2 50 " Moulder's Text-book 12mo, 2 50 Wiechmauu's Sugar Analysis 8vo, 2 50 Woodbury's Fire Protection of Mills 8vo, 2 50 MATERIALS OF ENGINEERING. STRENGTH ELASTICITY RESISTANCE, ETC. (See also ENGINEERING, p. 6.) Baker's Masonry Construction 8vo, .$5 00 Beardslee and Kent's Strength of Wrought Iron 8vo, 1 50 Bovey's Strength of Materials 8vo, 7 50 Burr's Elasticity and Resistance of Materials Svo, 5 00 Byrne's Highway Construction Svo, 5 00 Carpenter's Testing Machines and Methods of Testing Materials Church's Mechanic's of Engineering Solids and Fluids Svo, 6 00 Du Bois's Stresses in Framed Structures 4to, 10 00 Hatfield's Transverse Strains Svo, 5 00 Johnson's Materials of Construction Svo, 6 00 Lanza's Applied Mechanics 3vo, 7 50 " Strength of "Wooden Columns Svo, paper, 50 Merrill's Stones for Building and Decoration Svo, 5 00 Merriman's Mechanics of Materials Svo, 4 00 Strength of Materials 12mo, 1 00 Patton's Treatise on Foundations Svo, 5 00 Rockwell's Roads and Pavements in France 12mo, 1 25 Spalding's Roads and Pavements. 12mo, 2 00 Thurston's Materials of Construction , , Svo, 5 00 Thurston's Materials of Engineering 3 vols., Svo, 8 00 Vol. I., Non-metallic. Svo, 200 Vol. II. , Iron and Steel Svo, 3 50 Vol. III., Alloys, Brasses, and Bronzes Svo, 2 50 Weyrauch's Strength of Iron and Steel. (Du Bois.) Svo, 1 50 Wood's Resistance of Materials Svo, 2 00 MATHEMATICS. CALCULUS GEOMETRY TRIGONOMETRY, ETC. Baker's Elliptic Functions Svo, 1 50 Ballard's Pyramid Problem Svo, 1 50 Barnard's Pyramid Problem Svo, 1 50 Bass's Differential Calculus 12mo, 4 00 Brigg's Plane Analytical Geometry , 12mo, 1 00 Chapman's Theory of Equations 12mo, 1 50 10 Chessin's Elements of the Theory of Functions Compton's Logarithmic Computations 12nio, $1 50 Craig's Linear Differential Equations 8vo, 5 00 Davis' s Introduction to the Logic of Algebra 8vo, 1 50 Halsted's Elements of Geometry ..8vo, 1 75 Synthetic Geometry 8vo, 150 Johnson's Curve Tracing 12mo, 1 00 " Differential Equations Ordinary and Partial 8vo, 3 50 " Integral Calculus 12mo, 1 50 " " " Unabridged Least Squares , 12mo, 150 Ludlow's Logarithmic and Other Tables. (Bass.) 8vo, 2 00 Trigonometry with Tables. (*Bass.) 8vo, 300 Mahan's Descriptive Geometry (Stone Cutting) 8vo, 1 50 Merriman and Woodward's Higher Mathematics 8vo, 5 00 Merrimau's Method of Least Squares 8vo, 2 00 Parker's Quadrature of the Circle 8vo, 2 50 Rice and Johnson's Differential and Integral Calculus, 2 vols. in 1, 12mo, 2 50 Differential Calculus 8vo, 350 Abridgment of Differential Calculus.... 8vo, 150 Searles's Elements of Geometry 8vo, 1 50 Totten's Metrology 8vo, 2 50 Warren's Descriptive Geometry 2 vols., 8vo, 3 50 ' ' Drafting Instruments 12mo, 1 25 '< Free-hand Drawing 12mo, 100 " Higher Linear Perspective 8vo, 3 50 " Linear Perspective 12mo, 1 00 " Primary Geometry 12mo, 75 Plane Problems... 12mo, 125 " Plane Problems 12mo, 125 " Problems and Theorems 8vo, 2 50 Projection Drawing 12mo, 1 50 Wood's Co-ordinate Geometry 8vo, 2 00 Trigonometry 12mo, 100 Woolf's Descriptive Geometry Royal 8vo, 3 00 11 MECHANICS- MACHINERY. TEXT-BOOKS AND PRACTICAL WORKS. (See also ENGINEERING, p. 6.) Baldwin's Steam Heating for Buildings .' . . .12mo, $2 50 Benjamin's Wrinkles and Recipes 12mo, 2 00 Carpenter's Testing Machines and Methods of Testing Materials 8vo, Chordal's Letters to Mechanics 12mo, 2 00 Church's Mechanics of Engineering 8vc, 6 00 " Notes and Examples in Mechanics 8vo, 2 00 Crehore's Mechanics of the Girder 8vo, 5 00 Cromwell's Belts and Pulleys 12mo, 1 50 Toothed Gearing 12mo, 150 Compton's First Lessons in Metal Working 12mo, 1 50 Dana's Elementary Mechanics 12mo, 1 50 Dingey's Machinery Pattern Making 12mo, 2 00 Dredge's Trans. Exhibits Building, World Exposition, 4to, half morocco, 10 00 Du Bois's Mechanics. Vol. I., Kinematics 8vo, 3 50 Vol. II., Statics .8vo, 400 Vol. III., Kinetics 8vo, 3 50 Fitzgerald's Boston Machinist 18mo, 1 00 Flather's Dynamometers 12nio, 2 00 Rope Driving 12mo, 200 Hall's Car Lubrication 12mo, 1 00 Holly's Saw Filing 18mo, 75 Jones Machine Design. Part I, Kinematics 8vo, 1 50 " " Part II, Strength and Proportion of Machine Parts Lanza's Applied Mechanics 8vo, 7 50 MacCord's Kinematics .8vo, 5 00 Merriman's Mechanics of Materials .8vo, 4 00 Metcalfe's Cost of Manufactures .8vo, 5 00 Michie's Analytical Mechanics 8vo, 4 00 Mosely's Mechanical Engineering. (Mahan.) 8vo, 5 00 Richards's Compressed Air 12mo, 1 50 Robinson's Principles of Mechanism 8vo, 3 00 Smith's Press-working of Metals 8vo, 8 00 12 The Lathe and Its Uses . 8vo, $6 00 Thurstou's Friction and Lost Work 8vo, 3 00 The Animal as a Machine 12mo, 1 00 Warren's Machine Construction 2 vols., 8vo, 7 50 Weisbach's Hydraulics and Hydraulic Motors. (Du Bois.)..8vo, 500 Mechanics of Engineering. Vol. III., Part I., Sec. I. (Klein.)... 8vo, 500 Weisbach's Mechanics of Engineering. Vol. III., Part I., Sec. II (Klein.) 8vo, 500 Weisbach's Steam Engines. (Du Bois.) 8vo, 5 00 Wood's Analytical Mechanics 8vo, 3 00 " Elementary Mechanics 12mo, 125 Supplement and Key 1 25 METALLURGY. IKON GOLD SILVER ALLOYS, ETC. Allen's Tables for Iron Analysis 8vo, 3 00 Egleston's Gold and Mercury 8vo, 7 50 Metallurgy of Silver 8vo, 750 * Kerl's Metallurgy Copper and Iron 8vo, 15 00 * " Steel, Fuel, etc 8vo, 1500 Kunhardt's Ore Dressing in Europe. 8vo, 1 50 Metcalf Steel A Manual for Steel Users 12mo, 200 O'Driscoll's Treatment of Gold Ores. 8vo, 2 00 Thurston's Iron and Steel 8vo, 3 50 Alloys 8vo, 250 Wilson's Cyanide Processes 12mo, 1 50 MINERALOGY AND MINING. MINE ACCIDENTS VENTILATION ORE .DRESSING, ETC. Barriuger's Minerals of Commercial Value. . . .oblong morocco, 2 50 Beard's Ventilation of Mines 12mo, 2 50 Boyd's Resources of South Western Virginia 8vo, 3 00 '": " Map of South Western Virginia Pocket-book form, 2 00 Brush and Penfield's Determinative Mineralogy 8vo, 3 50 Chester's Catalogue of Minerals 8vo, 1 25 " " " " paper, 50 Dictionary of the Names of Minerals 8vo, 3 00 Dana's American Localities of Minerals , 8vo, 1 00 13 Dana's Descriptive Mineralogy. (E. S.) . . . .8vo, half morocco, $12 50 Mineralogy and Petrography. (J. D.) 12mo, 200 " Minerals and How to Study Them. (E. S.). 12mo, 1 50 " Text-book of Mineralogy. (E. S.) 8vo, 3 50 ^Drinker's Tunnelling, Explosives, Compounds, and Rock Drills. 4to, half morocco, 25 00 Egleston's Catalogue of Minerals and Synonyms 8vo, 2 50 Eissler's Explosives Nitroglycerine and Dynamite 8vo, 4 00 Goodyear's Coal Mines of the Western Coast . 12mo, 2 50 Hussak's Rock forming Minerals. (Smith.) Svo, 2 00 Ihlseng's Manual of Mining. . Svo, 4 00 Kuuhardt's Ore Dressing in Europe Svo, 1 50 O'Driscoll's Treatment of Gold Ores Svo, 2 00 Rosenbusch's Microscopical Physiography of Minerals and Rocks. (Iddiugs.) Svo, 500 Sawyer's Accidents in Mines Svo, 7 00 StDskbridge's Rocks and Soils Svo, 2 50 Walke's Lectures on Explosives Svo, 4 00 Williams's Lithology Svo, 3 00 Wilson's Mine Ventilation 16nio, 1 25 STEAM AND ELECTRICAL ENGINES, BOILERS, Etc. STATION AH Y MARINE LOCOMOTIVE GAS ENGINES, ETC. (See also ENGINEERING, p. 6.) Baldwin's Steam Heating for Buildings. 12mo, 2 50 Clerk's Gas Engine , 12mo, 4 00 Ford's Boiler Making for Boiler Makers ISmo, 1 00 Hemeuway's Indicator Practice 12mo, 2? 00 Hoadley's Warm-blast Furnace Svo, 1 50 Kneass's Practice and Theory of the Injector Svo, 1 50 MacCord's Slide Valve Svo, 2 00 *Maw's Marine Engines Folio, half morocco, 18 00 Meyer's Modern Locomotive Construction 4to, 10 00 Peabody and Miller's Steam Boilers Svo, 4 00 Peabody's Tables of Saturated Steam Svo, 1 00 Thermodynamics of the Steam Engine Svo, 5 00 Valve Gears for the Steam Engine Svo, 2 50 Pray's Twenty Years with the Indicator Royal Svo, 2 50 Pupin and Osterberg's Thermodynamics 12nio, 1 25 Reagan's Steam and Electrical Locomotives 12mo, 2 00 Routgen's Thermodynamics. (Du Bois.) Svo, 5 00 Sinclair's Locomotive Running 12mo, 2 00 Thurston's Boiler Explosion 12mo, 1 50 14 Thurstoii's Eugine and Boiler Trials 8vo, $5 00 Manual of the Steam Engine. Part I., Structure and Theory, 8vo, 7 50 Manual of the Steam Engine. Part II., Design, Construction, and Operation 8vo, 7 50 2 parts, 12 00 Philosophy of the Steam Engine 12ino, 75 Reflection on the Motive Power of Heat. (Caruot.) 12mo, 1 50 " Stationary Steam Engines 12mo.. 1 50 Steam-boiler Construction and Operation 8vo, 5 00 Spaugler's Valve Gears 8vo, 2 50 Trowbridge's Stationary Steam Engines 4to, boards, 2 50 Weisbach's Steam Engine. (Du Bois.) 8vo, 5 00 Whitham's Constructive Steam Engineering 8vo, 10 00 Steam-engine Design 8vo, 6 00 Wilson's Steam Boilers. (Flather.) 12mo, 2 50 Wood's Thermodynamics, Heat Motors, etc 8vo, 4 00 TABLES, WEIGHTS, AND MEASURES. FOR ACTUARIES, CHEMISTS, ENGINEERS, MECHANICS METRIC TABLES, ETC. Adriauce's Laboratory Calculations 12rno, 1 25 Allen's Tables for Iron Analysis., 8vo, 3 00 Bixby's Graphical Computing Tables Sheet, 25 Comptou's Logarithms 12mo, 1 50 Craudall's Railway and Earthwork Tables 8vo, 1 50 Egleston's Weights and Measures 18mo, 75 Fisher's Table of Cubic Yards Cardboard, 25 Hudson's Excavation Tables. Vol. II 8vo, 1 00 Johnson's Stadia and Earthwork Tables 8vo, 1 25 Ludlow's Logarithmic and Other Tables. (Bass.) 12mo, 2 00 Thurstoii's Conversion Tables Svo, 1 00 Totteu's Metrology Svo, 2 50 VENTILATION. STEAM HEATING HOUSE INSPECTION MINE VENTILATION. Baldwin's Steam Heating 12mo, 2 50 Beard's Ventilation of Mines 12mo, 2 50 Carpenter's Heating and Ventilating of Buildings. Svo, 3 00 Gerhard's Sanitary House Inspection Square 16ino, 1 00 Mott's The Air We Breathe, and Ventilation 16mo, 1 00 Reid's Ventilation of American Dwellings 12mo, 1 50 Wilson's Mine Ventilation 16mo, 1 25 15 niSQELLANEOUS PUBLICATIONS. Alcott's Gems, Sentiment, Language Gilt edges, $5 00 Bailey's The New Tale of a Tub 8vo, 75 Ballard's Solution of the Pyramid Problem 8vo, 1 50 Barnard's The Metrological System of the Great Pyramid. .8vo, 1 50 Davis' Elements of Law 8vo, 2 00 Emmon's Geological Guide-book of the Rocky Mountains. .8vo, 1 50 Ferrel's Treatise "on the Winds 8vo, 4 00 Haines' Addresses Delivered before the Am. Ry. Assn. . . .12mo. 2 50 Mott's The Fallacy of the Present Theory of Sound . . Sq. 16mo, 1 00 Perkins's Cornell University Oblong 4to, 1 50 Ricketts's History of Rensselaer Polytechnic Institute. . . . 8vo, 3 00 Rctherham's The New Testament Critically Emphasized. 12m o, 1 50 " The Emphasized New Test. A new translation. large 8vo, 2 00 Totteu's An Important Question in Metrology. 8vo, 2 50 Whitehouse's Lake Mceris Paper, 25 * Wiley's Yosemite, Alaska, and Yellowstone 4to, 3 00 HEBREW AND CHALDEE TEXT=BOOKS. FOR SCHOOLS AND THEOLOGICAL SEMINARIES. Gesenius's Hebrew and Chaldee Lexicon to Old Testament. (Tregelles.). Small 4to, half morocco, 5 00 Green's Elementary Hebrew Grammar. 12mo, 1 25 " Grammar of the Hebrew Language (New Edition). 8 vo, 3 00 " Hebrew Chrestomathy. 8vo, 2 00 Letteris's Hebrew Bible (Massoretic Notes in English). 8vo, arabesque, 2 25 Luzzato's Grammar of the Biblical Chaldaic Language and the Talmud Babli Idioms 12mo, 1 50 MEDICAL. Bull's Maternal Management in Health and Disease 12mo, 1 00 Hammarsteu's Physiological Chemistry. (Mandel.) 8vo, 4 00 Mott's Composition, Digestibility, and Nutritive Value of Food. Large mounted chart, 1 25 Ruddiman's Incompatibilities in Prescriptions. 8vo, 2 00 Steel's Treatise on the Diseases of the Ox. ... 8vo, 6 00 " Treatise on the Diseases of the Dog 8vo, 3 50 Worcester's Small Hospitals Establishment and Maintenance, including Atkinson's Suggestions for Hospital Archi- tecture i i ia/ 1 25 16 UNIVERSITY OF CALIFORNIA LIBRARY THIS BOOK IS DUE ON THE LAST DATE STAMPED BELOW OCT 28 1916 AY SI > r ' 8 HA.. 4 1920 DEC 1 1988 /If/jy APR 2 7 1966 Q -61-85 30m-l,'15