MOUNT WILSON OBSERVATORY. 235 



found to be GO and KO, respectively, the former being a giant and the latter 

 probably a dwarf. The brighter stars are of types B8 and F3. A prelimi- 

 nary orbit has been computed for RS Canum Venaticorum. 



(10) Examination of high-dispersion spectrograms has brought out the 

 fact that the intensity of the zinc triplet XX4680, 4722, and 4810 is relatively 

 greatest in Sirius and falls off in Procyon and Arcturus. This behavior con- 

 forms to the ionization theory of Saha, and was predicted by Russell from his 

 study of sun-spot spectra. 



(11) Measurements by Mr. Sanford of stellar spectrograms taken with a 

 dispersion of three prisms and a 102-cm. camera indicate that on photographs 

 of this scale the differences found by Adams between the velocities given by 

 arc and enhanced lines may be observed without difficulty. 



(12) Mr. Merrill has begun a general physical investigation of red stars, 

 with especial reference to those of the new spectral type S. These stars 

 appear to form a third branch of the spectral sequence cognate with the K-M 

 and R-N branches. 



(13) From a study based on (a) spectral lines and bands, (6) relative strength 

 of various portions of the continuous spectrum, (c) heat indices as observed by 

 Nicholson and Pettit, Mr. Merrill concludes that the effective temperature 

 of the reversing layer is lower in variables of type M than in those of other 

 spectral types. 



(14) Mr. Humason has discovered 7 new B-type stars with bright hydrogen 

 lines during the year. A total of 74 such stars has now been found at Mount 

 Wilson. In the course of this work in the spectral region near Ha, 2 F-type 

 and 2 M-type stars with bright hydrogen lines have been discovered. 



(15) Numerous measures have been made by Mr. Merrill and Mr. Humason 

 of the complicated spectrum of B. D. + ll 4623, a star which showed recently 

 a remarkable outburst of helium emission. A second star, H. D. 161114, 

 has been found to show a similar spectrum. 



MASSES AND DENSITIES OF STARS. 



Mr. Seares has finished his investigation of the masses of stars referred to 

 in the last report. By combining the mean masses of the dwarf stars with the 

 mean values of their space-velocities, as derived by Messrs. Adams, Stromberg 

 and Joy, he has found that for types F0 and later the kinetic energy of transla- 

 tion of these stars is nearly constant, notwithstanding the occurrence of 

 stream motion among them. With the aid of Professor Campbell's radial 

 velocities, the comparison was extended to the B and A stars. The kinetic 

 energy of the B's seems to be less than that for the other types, but from A0 

 through to Ma, the deviations from the mean energy constant are accidental 

 and of the order of the inherent uncertainty of the data. 



The extent to which this result holds for the giant stars is open to question. 

 In spite of this uncertainty, an attempt to derive values of the masses of the 

 giants was made by combining their space-velocities with the mean energy 

 constant. These results, together with the masses of the dwarf stars, gave a 

 series of distribution curves showing approximately the relation of mean 

 mass to absolute magnitude and spectral type for all classes of stars. 



To obtain corresponding values of the mean density, the question of surface 

 brightness was re-examined, attention being given to the differences between 



