MOUNT WILSON OBSERVATORY. 241 



greater than that of the Galaxy, and for a considerable number of the 

 brighter spirals the factor is of the order of 100. 



If we suppose the spirals to consist of aggregations of stars similar to 

 our own system, we arrive at the following relation between the number 

 of stars per unit volume, 5, and the linear dimensions, T, 



The subscripts S and G refer to the spirals and the galactic system, 

 respectively, and I is the ratio of the surface brightness of the nebulae 

 to that of the Galaxy. As remarked above, this quantity is always 

 greater than unity, and in a considerable number of cases attains values 

 of the order of 100. Alternative conclusions concerning the relative 

 values of 8 and T are obvious, and it appears that if we arrange the 

 spirals and the galactic system according to one or the other of these 

 characteristics, the Galaxy will be at one end, if not actually out- 

 side, the series thus formed. If, therefore, the spirals really consist of 

 stars, they are not in general comparable with our own system in respect 

 to both stellar density and linear dimensions. 



Observations of Color. 



The measurement of the color of stars of known intrinsic brightness 

 by the method of exposure ratios has been continued by Mr. Scares. 

 In addition to the blue-yellow ratio, a considerable number of measures 

 has been made in the red in combination with other spectral regions in 

 the hope of discovering criteria which would permit the determination 

 of both spectrum and luminosity from observations of color alone. For 

 the regions thus far used the problem seems to be indeterminate, owing 

 to the fact that the effect of luminosity on color enters in such a manner 

 that it can not be separated from that due to spectral type. If the 

 spectrum is known, observations of color, at least in the case of the late 

 A's and the G and K stars, can be made with sufficient precision to 

 afford approximate values of the luminosity, and hence of the distance. 

 But in that case there is no great gain, for the distance can then be 

 obtained directly from the spectrum. 



Intimately related to this question is the dependence upon lumin- 

 osity of the effective temperatures calculated by Planck's radiation 

 formula. We should expect a dwarf star to yield a higher black-body 

 temperature than a giant of the same spectral type, and an examina- 

 tion of the measures of Wilsing, Scheiner, and Miinch shows this to be 

 the case. The behavior of the spectral lines used as criteria for the 

 determination of absolute magnitude would indicate the contrary, 

 however, which throws some light upon the deviations of black-body 

 temperatures from the true temperatures of stars. 



