270 CARNEGIE INSTITUTION OF WASHINGTON. 



stars probably belong to the group, but as yet onl}' single observations 

 are available for them. The average spectral type of the stars found 

 to belong to the group is B 6. 



(4) The radial velocities of the Andromeda nebula and the bright line 

 nebula N. G. C. 1068 have been determined from photographs taken by 

 Mr. Pease with a small spectrograph at the primary focus of the 60-inch 

 reflector. They are as follows: 



Andromeda nebula —329 km. 



N. G. C. 1068 +765 



Absolute Magnitude Effects in Stellar Spectra. 



Photographs of the spectra of pairs of stars taken during the year 

 confirm the conclusion referred to in the last annual report that, of 

 stars having the same types of spectra, the more distant stars are 

 fainter in the violet portion of the continuous spectrum. The prob- 

 ability that at least a part of this effect is due to the greater intrinsic 

 brightness of the more distant stars has been increased to some extent 

 by recent observations. On account of uncertainties in parallax deter- 

 minations, it is difficult to select suitable pairs of stars of known dis- 

 tance. Accordingly the question is now being investigated with the 

 aid of observations on two lists of selected stars: 



(1) Faint and bright stars in the Hyades which are known to belong 

 to the Taurus cluster. Since these stars are at nearly the same dis- 

 tance, a difference in the violet portion of the continuous spectrum for 

 stars of the same spectral type must be ascribed to the effect of intrinsic 

 brightness. The exceptionally unfavorable observing weather during 

 the winter made it possible to obtain only a few photographs of these 

 stars. So far as the evidence extends it indicates that the fainter stars 

 are relatively stronger in the violet part of the spectrum. 



(2) The observing list of small proper-motion stars with magnitudes 

 between 7.5 and 9 affords the opportunity of making a comparison of 

 the intensity of the spectrum of these stars with those of stars of the 

 same average proper motion but about 2.5 magnitudes brighter, which 

 have already been observed for radial velocity. In the average the two 

 groups of stars should have closely the same distance, and the fainter 

 stars apparently should also be the fainter stars intrinsically. The 

 results so far obtained indicate that the fainter stars are stronger in 

 the violet portion of the spectrum. 



The method of using the relative intensities of certain spectral lines 

 as a criterion of absolute magnitude, which was developed mainly by 

 Mr. Kohlschiitter, and described in the last annual report, has been 

 continued during the year. The results found fully confirm those 

 obtained previously, and appear to warrant the use of this means for 

 the actual determination of individual parallaxes. At present the chief 

 need is a more accurate standard of reference for the very distant stars 

 which, in the absence of parallax measures, has been based upon the 



