By common consent the nebula is regarded as being the primordial matter. The first stage of development is represented, according to Huggins ((Atlas of Representative Stellar Spectra,' London 1899), a masterpiece with superb illustrations, by a star like A of Orionis, one of the stars of the great nebula. Orion and helium stars are followed in devel opment by the white stars of Secchi's first type. Evolution can be traced step by step, through stages like those represented by the stars a Lyra, Sirius, Castor (fainter star), a Aquila, Procyon, Y Cygni, till we come to the fully developed second type star like Capella and the sun. Increased absorption at the violet end of the spectrum gives the red stars of Secchi's third and fourth types which, according to Hug gins, develop in parallel lines. The period of in creasing old age is evident from the carbon ab sorption bands, and it is easy to imagine this ab sorption increasing in amount till the whole light of the star is cut off. But even at this stage, when the star gives no light, the spectroscope is not powerless to follow, for if the dark star accompanies a bright one, its presence is re vealed through a change in the motion of the line of sight. There has been quite consider able discussion as to which star, Sirius or the sun, is in the hotter stage of development. The color of Sirius, and the maximum in its spec trum being more toward the violet end than in the case of the sun, would seem to indicate a higher temperature. Huggins and a great many astronomers think that the sun is in the hottest state, but that the great absorption in its atmosphere, compared with that of Sirius, makes the color of the sun yellow.
All classifications of stellar spectra begin with white stars and end with red. If we in terpret this as the result of a change in tem perature corresponding to a progress in evolu tion,. as shown by Wien's law, and if the in tensities of the different regions of the spec trum are measured by a spectral photometer and Planck's energy equation is applied, as in the case of the sun, then we find the following approximate effective temperatures of the stars of the different classes: Class 0e5, 12,000° C; classes B and A, C; class F, C; class G C; class K, 4,000° C; and class M, C. Other methods of evolution among the stars have been shown possible by Lock yer, Russell and others.
Motions of the Stars.— Proper motion is determined from meridian circle observations. The motions are measured in angular units of seconds of arc per year. The most complete work is the 'Preliminary General of Boss containing the proper motions of 6,1:: stars. Radial velocities are measured by the spectrograph in linear units of miles, or kilo metres per second. The most prominent work ers in this field have been: Vogel, Duner, Campbell, Frost, Belopolsky, Hartmann, Schles inger, Plaskett and Adams, who together have pretty thoroughly surveyed the northern hemi sphere. Work in the southern hemisphere has been done by the Lick Observatory station at Santiago, Chile, and by the Cape Observatory in South Africa. In stellar spectroscopic work the advantage is shown of the great light-gathering power of the large modern reflecting telescopes. The advantage of the reflector over the ref rac tor in this field of work is that there is no chromatic aberration, so the light of all wave lengths is brought to a focus on the slit. The small field of good definition, which the reflec tor shows in comparison with the refractor, is no drawback in spectroscopic work. With the
five-foot reflector of Mount Wilson, it has been possible to photograph the spectra of stars as faint as the 12th magnitude, while with the 100 inch reflector, it will be possible to photograph the spectra of stars of the 15th magnitude, or even fainter. In addition to measuring the motion of the star toward, or away from, the observer in the line of sight, the spectrograph has permitted the discovery of an entirely new class of stars known as spectroscopic binaries. The first discovered was Mizar by Pickering, who noticed the periodic doubling of the lines of the spectrum. It is comparatively rare that both stars are of so nearly equal bright ness that both show the stellar absorption lines. But when the motion of the star in the line of sight varies in amount (after eliminating the motion of the earth about the sun), this change in velocity is certain proof of the presence of a second body. Repeated observations of such stars permit the determination of their orbits. It is found that the periods have a wide range, from 0.19 days in the case of B Cephei, to 6.0 years as in a Orionis, this latter period being longer than that of the shortest period of a visual binary d Equulei, which has a period of 5.7 years. Among the brighter stars, about one star in every four is found to be a spec troscopic binary.
The most complete work on radial velocities so far published is Motions,' Camp bell, Yale University Press, 1913. Campbell has made the remarkable discovery that as the stars increase in age their velocities increase ac cording to the following table: Spectral Classes Number of Average radial Stan Velocities 0 and B 141 S.pp km.
A 533 9.94 F 159 15.90 G and S ts.ls M 72 16 SS There is no satisfactory explanation of this curious increase of speed with age unless we accept that of Adams (Astrophysical Journal, XLII, 172, 1915). The early type stars have smaller proper motions and are at much greater distances from us than the later type stars. Due to the fact that Campbell's work has been lim ited to stars brighter than 5.0 magnitude, he has, in consequence, measured stars of higher average velocities in the later types than in the earlier. Kapteyn has shown that there is an intimate connection between proper motion, radial velocity, magnitude and spectral type of the stars, and has derived empirical formulae showing the connection. The most complete work in this field has been done by van Rhijn (Astrophysical Journal, XLIII, 36, 1916). This work shows very clearly that the early type stars belonging to the 0, B and A classes on the average are mush more distant than the stars belonging to the later types. Adams at Mount Wilson has gone a step farther, and has made the wonderful discovery ((Proceedings' of the National Academy, April 1916), that the relative intensities of certain lines in stellar spectra show changes depending on the absolute luminosity of the stars, and by means of meas ures of these intensities, Adams has been able to determine the parallaxes of the stars. The spectrum of the sun is almost identical with that of Capella, which shows that the sun is a star, rendered brighter and bigger on account of its nearness. The stellar magnitude of the sun is —26.4 on the. same scale that Sirius is a star of magnitude-1.4. Thus, according to Newcomb ((Stars,' p. 27), the sun gives us: 10,000,000,000 times the light of Sirius.