Wave-lengths.— The unit generally used in the measurement of wave-lengths is the Angstrom, or tenth-metre, or metre divided by 10'°. A great change has recently come over our ideas of the wave-length. Formerly the wave-length was looked upon as an in variable property of a line in the spectrum, unalterably fixed by nature, and consequently it was thought that a wave-length determina tion would give a standard measure of dis tance more reliable even than that obtained by the use of the International Metre (Mi chelson, 'Astronomy and Astrophysics,' XIII, 92, 1894). But we now know that the position of the lines in the spectrum may vary with the pressure of the gas in which they are produced, and moreover, single lines, by the action of a magnetic field, have been sep arated into as many as 19 different components. The shift in the lines of the spectrum due to motion in •the line of sight, which has been shown experimentally in the laboratory, has given rise to many interesting developments in astrophysics, the discovery of an entirely new class of stars called spectroscopic binaries, the measurement of the axial rotation of the sun and Jupiter, and has confirmed in a magnificent manner the meteoric constitution of Saturn's rings.
Spectra of the Elements.— The invention of the concave grating and the manufacture of nearly perfect gratings, plane and concave, by Rowland, enormously increased the accuracy of the wave-lengths in the spectra of the ele ments. The chief investigations in this field have been carried on by Rowland and his as sistants, by Kayser, Runge, Paschen, Hassel berg, Liveing and Dewar, Eder and Valenta, Exner and Haschek, Hartley and Adeney, Trowbridge, Ames, Lockyer, Deslandres, Lohse and others. The most complete tables of wave lengths are those of Exner and Haschek, in which the wave-lengths are given to 0:01 Angs trom. Other tables are Watt's 'Index of Spectra.' The standards on which all wave length tables are based are Rowland's
curate wave-lengths have been determined by Lyman of Harvard, using a concave grating in a vacuum and photographic plates poor in gela tine, as shown by Schumann. Lyman has reached as short wave-lengths as 600 (As trophysical Journal, XLIII, 89, 1916), and the limit of short wave-lengths can probably be pushed still farther by this same method.
Line Series.— That there is some orderly arrangement in the lines of a spectrum was shown by Balmer's law for the hydrogen se ries in 1885, and by the presence of numerous triplets in the spectra of magnesium, calcium and zinc. Researches by Kayser and Runge, Rydberg and Schuster, with the development by them of empirical and mathematical for mulae, have led to a great deal of interesting information regarding these series of lines. Our knowledge of the spectrum of hydrogen is derived mainly from the observations of the heavenly bodies. In the laboratory, compara tively few lines are found in the series obey ing Balmer's law, but these lines are well shown in the spectra of hydrogen-type or class A stars, and as many as 32 lines of this series have been measured by Mitchell in the eclipse spectra of the sun. The further interdependence of astronomy and physics was clearly demon strated when Pickering in 1896 discovered a new series of lines in the spectrum of the star e Puppis. It was not till 1892 that Fowler of London succeeded in finding in the laboratory some lines of the Puppis series, and also A 4686, and other lines of the Principal series of hydrogen. At the same time Fowler discov ered a second Principal series of hydrogen, and comes to the conclusion that
are no spe cial kinds of matter in celestial bodies and that most, if not all, of the celestial spectra are well within range of laboratory experiments.* (aMonthly Notices," R. A. S., LXXIII, 62, 1912). However, in a great majority of the elements the series already discovered com prise only a small percentage of the total num ber of lines. The exact meaning of these se ries is as yet unknown, although several very promising attempts have been made to explain them from theoretical considerations. The chief among them may be mentioned those of Julius, Ames, Kiivesligethy and Stoney. Stoney
Roy. Soc.,) Dublin 1891) has sought to explain multiple lines from dynamical con siderations, comparing the motions of the mole cule with those of the bodies of the solar sys tem whose motions in ellipses are perturbed by the presence of other bodies. Stoney, more over, shows that the conclusions drawn by these dynamical methods may also be considered valid under the electromagnetic theory of light, a statement which receives support from Pres ton's observations of the Zeeman effect (Philo sophical Magazine, XLVII, 176, 1899). For de tailed information on spectral series consult Kayser's (Handbuch der Spectroscopic' Vol. II.