in Ottawa in showing that the calcium lines do not share in the same variation in amplitude as the other lines in the spectrum. On August 27 there is scarcely any range indicated. On September 2 the range is not more than twenty-five kilometers, while the other lines show a range of seventy. When the amplitude of the hydrogen and helium spectral lines is large, the calcium lines show a measurable smaller amplitude. When the hydrogen and helium lines are oscillating through a small range, the amplitude of the K and H lines is either zero or masked by the errors of measurement. The velocity of the centre of mass of the binary system, as determined from the calcium lines, is in substantial agreement with that determined from the other lines. This holds for cach year's observation. Thus we have: The wave length of the calcium lines if increased slightly would remove the systematic difference. The point that is significant is that the calcium lines strengthen the result, if indeed that needs strengthening, that the velocity of the centre of mass has varied since 1913. Moreover, this variation shows in the clearest way that the calcium vapors causing the absorption are moving with the star. One other point is worth recording from this investigation. The calcium lines show very little variation of character such as is shown by the other spectral lines. Whether we regard stars of this class as true binaries, or whether we ascribe the variation in the observed radial velocity to some pulsation of a single star, as suggested by Shapley, the calcium vapor as a huge cloud surrounding the condensation or condensations in the centre is not greatly disturbed by the revolution within, if the star is a binary, or by the pulsations of the core if it be but a single star. On the hypothesis that there are two condensations and that the cloud rotates around their common centre of gravity, with a period "Contr. Mt. Wilson Sol. Obs., No. 92. coincident with the revolution of the stars in the centre, it is possible to explain1 very nicely both the cases of apparently stationary H and K and those cases where these lines share in part of the variation of the other lines in the spectrum by assuming different degrees of condensation for the calcium vapor. The phenomena can, however, be explained equally well by the pulsation theory, assuming that the absorption which gives rise to the calcium lines. occurs at a very high level and that the outer layers of the calcium cloud are not greatly disturbed by the pulsation taking place in the core in the centre. Summary.-1. From the measures of spectra of 12 Lacertae taken in 1918 it is definitely shown that the amplitude of the radial velocity curve varies from night to night, being as small as fifteen or twenty kilometers at some times and as large as seventy at others. 2. The shape of the curve is not constant, but a simple period will harmonize fairly satisfactorily the observations taken between 1908 and 1918, and serves to predict the approximate times of maximum and minimum velocities. 3. The spectral lines undergo a periodic variation in width, the lines being wide and diffuse at periastron passage and sharper and narrower at apastron. When the amplitude exhibited by the curve is small, the lines are generally more diffuse than when the range is large. 4. The velocity of the centre of gravity as determined from plates taken in any one year is constant, but plates taken in 1913, 1914, 1915 and 1918 show that it is subject to a long-period variation. 5. The amplitude exhibited by the calcium lines is smaller than for the other spectral lines, and they are always sharp and narrow even when the others have become wide and diffuse. When the range in the hydrogen and helium lines is high, then the amplitude exhibited by the calcium lines is easily discernible; when the hydrogen and helium lines show a small range, the This JOURNAL, Sept., 1916, p. 372. amplitude exhibited by the calcium is masked more or less compietely by the error of determination. 6. The velocity of the centre of mass, as determined from the calcium lines, shows the same variation from year to year as do the other lines in the spectrum. This shows very plainly that the calcium causing the absorption is moving with the star and that therefore it is probably in the form of a large cloud surrounding it. DOMINION ASTROPHYSICAL OBSERVATORY, VICTORIA, Dec. 4, 1918. THE SPECTROSCOPIC BINARY BOSS 1275 BY J. B. CANNON OSS 1275 (a = 5h 16m0, 8 + 29° 29', Spectral Type A, BOSS = Mag. 57) was discovered to be a spectroscopic binary at Mt. Wilson, and the measures of six plates were sent by Dr. Adams to this Observatory in October, 1916. Of these, two plates showed the components separated, and the other four the spectra superposed giving one measure. Forty plates were taken here and the following elements were derived from the measures of them : On the curve the double circles are those normal places common to the two components. The values of K, and Kare practically the same and the stars are hence about equal in mass. The lines where separated also show that there is little difference in brightness. DOMINION ASTRONOMICAL OBSERVATORY, OTTAWA, CANADA, |