ASTRONOMY: F. H. SEARES 
29 
galactic latitude —3°. With the addition of the data for Messier 11 
and for the four neighboring fields, it appears that the redness of the 
faint stars will be found to depend, as might be expected, upon galactic 
latitude, and in the mid-galactic regions will vary with the density of 
the star clouds. 
6. The presence of negative color indices for faint stars in the three 
widely separated galactic regions mentioned above shows that, if these 
stars are typical in absolute brightness, the dimensions of the galactic 
system in the plane of the Milky Way are many times greater than has 
been inferred from studies of variables and investigations of the motions 
and magnitudes of the brighter stars. 
1 For instance, see Parkhurst, Yerkes Actinometry, Astropk. Chicago, III., 36, 1912, 
(218, 225) and figure 5 on p. 56 of ML Wilson Contrib., No. 116. 
^Berkeley, Univ. Cal. Pub., Lick Obs. 11, 1913, Plate 62. 
» ML Wilson Contrib., No. 116, 1915, (81 ff.). 
♦These Proceedings, 2, 1916, (12-15). 
^Astroph. J., Chicago, III., 39, 1914, (361-369); [ML Wilson Contrib., No. 81]. 
Ubid., 42, 1915, (92-119); [ML Wilson Contrib., No. 100]. 
Ubid., 42, 1915, (120-132); [ML Wilson Contrib., No. 102]. 
8 Ml Wilson Contrib., No. 117, 1916. The unpublished results for the fields near Messier 
13 are provisional. 
' In N. G. C. 1647 Hertzsprung {loc. ciL) finds one star of photographic magnitude 12.40 
that is apparently of spectral type B. 
^°Astroph. J., Chicago, III., 42, 1915 (148-162); [ML Wilson Contrib., No. 104]. 
THE COLOR OF THE STANDARD POLAR STARS DETERMINED BY 
THE METHOD OF EXPOSURE-RATIOS 
By Frederick H. Scares 
MOUNT WILSON SOLAR OBSERVATORY, CARNEGIE INSTITUTION OF WASHINGTON 
Communicated by G. E. Hale, December 18, 1916 
An earlier note in these Proceedings describes a method of measuring 
the color of a star which depends upon the ratio of the exposure times 
necessary for its blue and yellow hght to produce images of the same 
size.^ A comparison of the observed exposure-ratio with a curve derived 
by combining similar ratios for stars of known color affords a means of 
expressing the results in terms of color-index or color-class. 
The method is expeditious, and, under favorable conditions, precise; 
it is entirely independent of stellar magnitudes, and hence avoids the 
systematic errors which so easily enter as a result of uncertainties in 
the magnitude scales or in their zero points. Moreover, the method is 
direct, in the sense that color is measured and not inferred from obser- 
vations of spectral type. The results thus include that part of the color 
