522 



SCIENCE 



[N. S. Vol. XXXIV. No. 877 



rigidly with a metal plate nearly in the focal 

 plane and that this combination shall remain fixed. 

 Eigidity is not required in any other parts of the 

 apparatus. The plate has two holes illuminated 

 by incandescent lamps which when lighted form 

 minute points on the photograph. A circle about 

 two centimeters in diameter is silvered in the cen- 

 ter of the objective, whose aperture is reduced to 

 this amount. Several exposures on the sun are 

 made at noon automatically by the clock, moving 

 the plate in declination after each exposure by an 

 amount equal to the diameter of the sun. At the 

 same time a current passes through the lamps 

 which thus impress two reference points upon the 

 photograph. At night the same or another plate 

 is exposed for a second or so to the stars, using the 

 full aperture, and the position of the trails with 

 regard to the reference points, compared with that 

 of the sun, gives the relative positions. 



Photographs were exhibited showing images of 

 the sun and stars upon the same plate. 



The Spectra of 76S Double Stars: Annie J. 



Cannon. 



A list was prepared of all stars in the ' ' General 

 Catalogue of Double Stars" by Burnham, and the 

 ' ' Eef erence Catalogue of Southern Double Stars ' ' 

 by Innes, in which the components are of magni- 

 tude 7.5, or brighter. A special examination of 

 the Harvard photographs was made to determine 

 the spectra of these stars. 



Classification of some Stellar Spectra Photo- 

 graphed with the Slit Spectroscope at the Alle- 

 gheny, LicTc and Terkes Observatories, Com- 

 pared with those talcen at Harvard with the 

 Objective Prism: Annie J. Cannon. 

 The Draper classification of stellar spectra de- 

 pends wholly on photographs taken with the ob- 

 jective prism. In view of the fact that a large 

 number of photographs is being made with the slit 

 spectroscope at various other observatories, it ap- 

 pears to be a matter of great importance to make 

 a comparative study of these spectra, and to deter- 

 mine whether the same system of classification will 

 apply to spectra obtained by these two widely 

 different methods. Accordingly, a preliminary 

 study has been made of 131 spectrograms, in- 

 cluding spectra from Class Od to Class Md. 



The Spectra and Colors of Bed Stars of Harvard 

 Classes N and E: J. A. Parkhurst. (Bead by 

 Dr. Slocum.) 

 In June, 1911, Professor E. C. Pickering sent 



the writer a list of ' ' Fourth Type Stars not Eed ' ' 



(Class E). This list was extended to include an 

 assortment of ordinary red stars (Class N). An 

 investigation of the spectra of these stars was 

 made from plates taken with the objective prism 

 on the 6-ineh Zeiss camera and with the Brashear 

 spectrograph on the 40-inch Yerkes refractor. 

 The ' ' Color-Index ' ' was found by comparing 

 the photographic magnitude with the so-called 

 ' ' visual ' ' magnitude obtained with color-sensitive 

 plates and the ' ' visual luminosity ' ' filter, on 

 plates taken with the Zeiss camera and the 2-foot 

 reflector. 



The paper, which will be published in full in 

 Astrophysical Journal, gives results for 17 stars, 

 for which the "color-indices" range from 1.11 to 

 5.60 magnitudes. The principal conclusions are: 



1. Excepting the first star, which apparently has 

 a composite spectrum, all are as red or redder 

 than a Tauri, and therefore seem to deserve the 

 name of ' ' red ' ' stars. 



2. No sharp line can be drawn between the 

 Classes N and E. 



The Orbits of the Spectroscopic Components of 



d Bootis: W. E. Harpee. 



Fifty-three spectrograms of this F-type star, 

 photographic magnitude 5.3, form the basis of the 

 determination of the elements of the orbit. The 

 plates were obtained mostly with the single-prism 

 instrument, but for considerably over half the 

 period of 9.60 days the spectra were separated 

 and measures were made on each component. 

 Elements were determined for each by the method 

 of least -squares, after preliminary elements had 

 been obtained graphically. These were in sub- 

 stantial agreement with each other; but a more 

 rigid determination was eifected by combining the 

 observation equations of the two components into 

 0716 set of normal equations, thereby deriving 

 uniform values for the elements. 



Studies of Bright Variable Stars: Joel Stebbins. 

 Observations with the selenium photometer have 

 been continued during the past year, and the 

 paper contains results for three stars, a Orionis, 

 8 Orionis and /3 Aurigw. a Orionis is a well-known 

 irregular variable, and the results for 1910-11 

 indicate a range of 0.2 mag. S Orionis has long 

 been suspected of variability, but it is now shown 

 to be an eclipsing variable with two minima, the 

 range being about 0.10 mag. The third star, 

 /3 Aurigce, has also been found to be an eclipsing 

 variable with a range of less than 0.10 mag. A 

 combination of the photometric and spectroscopic 

 results gives the actual dimensions of this binary 



