Appendix. 



Color-index and spectral type. 



22. The apparent magnitude of a star may be defined as the integrated 

 energy of the star as measured by a certain instrument. The instruments here 

 considered are the human eye and the photographic plate*. Eacl) instrument has 

 its special distributive function (compare L. M. N:o 67) and as to the nature of this 

 function the observed spectral distribution of the energy as well t\s the total energy, 

 h. e. the apparent magnitude, will vary from an instrument to another. As the 

 distributive function of the human eye may vary considerably from on individual 

 to another, it is convenient to substitute for the eye a photographic plate only pre- 

 j)ared (eventually completed by a filter) such as to produce the same distributive 

 function as that of the average human eye. Such magnitudes are called j)hotovisual. 



The difference between the photographic ** and the visual (or photovisual) 

 magnitude is called the color-index. The zero point of the two scales is according 

 to the proposal of the committee on magnitudes of the astrographic chart, chosen in 

 such a way that »for the stars of spectral type ^0 on the Harvard classification, 

 and of magnitude 5.5 to 6 B, the photographic and visual magnitudes shall be assu- 

 med equal, and the visual magnitudes shall be reckoned on the Harvard scale* 

 (Parkhurst, Aph. J. 36,170 (19 1 2)). 



Extensive series of observations for determining the color-indices of the star ■ 

 have been made at Göttingen, Harvai'd and Yerkes. 



23. The Yerkes observations have been performed by Parkhorst (»Yerkes Acti- 

 nometry» 1. c). It embraces all stars in B. D. brighter than the magnitude 7.5 situated 

 north of 8 = -)- 73°. All these stars have been observed photometrically at Potsdam 

 by Müller and Kempf (Publicationen des astrophysikalischen Observatoriums in 

 Potsdam, Bd XVII (1907)). Parkhurst does not confine himself to the use of these 

 visual magnitudes but determines independently with his camera the photovisual 

 magnitudes making use of a color filter and color-sensitive plates. Comparing his 



* There are in the last time even other instruments used witli great advantage in stellar 

 photometry. I think especially on the photo-electric methods used by Guthnick and others. 

 ** Even here it is necessary to use plates having an well defined distributive function. 



