272 CARNEGIE INSTITUTION OF WASHINGTON. 



star o- Eridani, which is of type AO, a result previously indicated by 

 the color-index determinations of other observers. 



Several attempts have been made to photograph the spectrum of 

 the companion of Sirius. Although no conclusive results have been 

 obtained, photographs secured under observing conditions which would 

 certainly show the spectrum of this star with the exposure times used 

 give no indication of the presence of any spectrum different in type 

 from that of Sirius. The hypothesis might at least be entertained that 

 the Ught from the companion is in part reflected hght and that this 

 may account for its low luminosity. Two direct photographs obtained 

 by iMr. van Maanen with and without the use of a color screen also 

 indicate no marked color index for the companion star. 



A photograph of the specti-um of Messier 15, made by Mr. Pease 

 with the small focal plane spectrograph, shows the spectra of about 18 

 stars ranging generally in type from A5 to F5, with one star about GO. 



With the same instrument Mr. Pease has also photographed the 

 spectrum of the Andromeda nebula, the total exposure time amounting 

 to 34 hours. The spectrum is of the G type, the more prominent lines 

 extending over a minute of arc on either side (E-W) of the nucleus. 

 As already stated, the mean velocit}^ from 9 lines measured by Mr. 

 Adams is —329 km., reduced to the sun. There seems to be no evi- 

 dence indicating either the presence of bright lines or of a rotational 

 displacement. 



The spectra of several planetarj^ nebulae were photographed by Mr. 

 Pease to determine the relative intensities of various lines with a view 

 to possible work with a Fabry-Perot etalon. 



PROFESSOR KAPTEYN'S INVESTIGATIONS. 



This 3'ear attention has been given mainl}^ to the three following 

 subjects: 



(1) The relations between the proper motions and the radial velocities 

 of stars of the spectral types F, G, K, andM (jointly with Mr. Adams). — 

 The question was mentioned in last year's report. Since then prelimi- 

 nary results have been published in the Proceedings of the National 

 Academy of Sciences. The distribution of the pecuUar velocities has 

 been investigated further, and results have been obtained from which 

 both the solar and the stream motions are completely eliminated. Thej^ 

 give a satisfactory explanation of the change of the mean radial velocity 

 with the proper motion. Traces were also found of a change of velocity 

 with absolute brightness, but as the evidence was very slender, a list of 

 stars has been drawn up for the purpose of settling the question. The 

 stars of this list are now under observation. 



(2) The determination of the average parallax of the stars of spectral 

 types N and 0. — Both of these classes of stars, but especially the lat- 

 ter, are very strongly condensed towards the Milky Way. It seems 



