34 



C. V. L. Charller 



absolute magnitudes. To whicli extent the magnitudes of the same spectral type 

 can differ is presently vaguely known, but we may get some idea of it from the 

 results regarding the brightest stars as well as from the parallax stars. I give in 

 the following table the values of M obtained for these stars. For the 0-stars the 

 results are obtained by Gyllenberg (L. M. 75), for the -ß-stars by myself (L. M. 

 II, 14), for the J.-stars by Malmqvist (L. M. 76), for the i^-stars by Lundahl (L. 

 M. 77), for the G- and Jf-stars by FIngk, for the ^-stars by Keantz and Lundahl. 

 The latter results are not yet published *. The absolute magnitudes of the parallax 

 stars are computed by Wicksell (L. M. II, 9, P. 101). 



TABLE 4. 

 Absolute magnitude. 



Sp. 



Bright stars 

 m < 6.0 



Parallax stars 



M 





M 



N 



0 

 B 

 A 

 F 

 G 

 K 



M , 



— 2.95 



— 1.10 

 + 0.81 



— 0.61 



— 1.03 



— 1.8Ü 



479 

 1251 



555 



41G 

 1142 



190 



- 2.31 

 + 1.47 

 + 3.07 

 + 3.85 

 + 4.10 

 + 8.17 



4 

 13 

 45 

 43 

 44 



9 



Mean 



— 1.02 



4033 



+ 3.53 



158 



Excepting the i?-stars we find that the parallax stars throughout have a fainter 

 absolute magnitude than the bright stars. The mean difference is 4,55. The diffe- 

 rence is largest for the ilf-stars and diminishes with increasing temperature of the 

 stars. Whether this fact can be generalized to embrace all stars of these types is 

 still an unsettled question. The theory of Russell for the evolution of the stars 

 goes in this direction but must still be regarded as only a hypothesis, though cer- 

 tainly as a very good working hypothesis. At any rate we must bear in mind 

 that these values of the absolute magnitudes are derived from a very limited range 

 of stars and that we must be prepared to find stars having as faint absolute magni- 

 tudes as, say, +7.0 certainly among the 6r-stars and perhaps also among stars of 

 the spectral types F or A. 



We have found in the preceding paragraph that the absolute magnitude of 

 the brightest stars in the globular clusters amounts, in the mean, to 7.1. Regarded 

 in the light of the table 4 we conclude that the stars in the globular clusters all 

 are to be considered as dwarf stars, less luminous than the majority of stars for 

 which till now the absolute magnitude has been determined. 



* In table 4 the stars of the types B% and B9) liave been excluded and moreover 16 K- 

 stars having large proper motions. 



