MOUNT WILSON SOLAR OBSERVATORY. 



223 



A continuation and extension of this investigation is now being car- 

 ried on by Mr. Stromberg. This will include a determination of the 

 solar motion from groups of stars of different absolute magnitude, the 

 relative effect of stream motion on stars of high and low luminosity, 

 and a study of the velocities of stars of different intrinsic brightness 

 with reference to the galactic plane. An interesting feature already 

 shown by the investigation is the marked condensation of the high- 

 luminosity stars of types F to M in the galactic plane, a result pre- 

 viously known to exist for the A and B stars. 



Spectroscopic Determinations of Luminosity and Parallax. 



The determination of the absolute magnitudes and the luminosities 

 of stars from a study of their spectra by the method described in the 

 report of last year has been continued bj^ Mr. Adams and Mr. Joy, 

 with the aid of the photographs of the stars observed for radial 

 velocity. Magnitude determinations are now available for about 800 

 such stars, many of them with measured trigonometrical parallaxes. 

 A Ust of 500 of these stars has been selected for publication; this 

 includes stars for which several photographs have been obtained and 

 a number of special interest, though with fewer observations. For 

 360 of these stars parallaxes have been determined in the usual way 

 by various observers. A comparison of the results, spectroscopic 

 minus directly measured values, is as follows : 



There are 59 stars in the hst with parallaxes measured by three or 

 more observers. The systematic difference in the case of these stars 

 is +0''001. In view of this comparison it is probably not too much 

 to state that the spectroscopic method, for all but the very largest 

 parallaxes, is capable of yielding results of nearly the same order of 

 accuracy as direct photographic methods with modern instruments. 



Two extensions of the method have been made. The first is the 

 use of certain additional spectral lines which has made it possible to 

 include stars of types A8 to F5 in the investigation. The second is an 

 attempt to derive criteria for the further subdivision of the dwarf 

 M-type stars. Hitherto all the known stars of this type have been 

 foimd to have nearly equal absolute magnitudes of about 10 to 11, but 



