ASTRONOMY: H. SHAPLEY 
435 
these three volumes is that most important for the present problem, 
since the spectra of all B stars (down to magnitude 8 or 8.5 and many 
fainter ones) from zero to nine hours in right ascension are included, 
and in the later hours of right ascension the stars of the Milky Way 
and those of the local cluster are much less clearly separated on the face 
of the sky. This difference in separation in- opposite hemispheres re- 
sults from the fact that the sun is some distance to one side of the central 
plane of the local cluster, thus giving a conspicuous Mip' to the encir- 
cling incHned belt of cluster stars, which minimizes the separation from 
the galactic equator in Scorpio and Ophiuchus and exaggerates it the 
most in Taurus and Orion. 
The amount of the apparent separation from the Milky Way for any 
given galactic longitude naturally depends on the average distance 
(brightness) of the cluster stars concerned. For those not more than 
200 parsecs distant from the sun, the maximum separation from the 
galactic equator is about 17° in Orion and Taurus, as indicated by the 
broken curve in all the accompanying figures. For stars at the distance 
of 400 parsecs, which appears to approximate the equatorial radius of 
the cluster, the maximum separation is a Httle less than 14 degrees, as 
shown by the lightly dotted curve in figure 2. 
Right ascension and decKnation are rectangular coordinates in figure 
1, which shows the course of the Milky Way throughout the first thir- 
teen hours of right ascension. The data are taken from Harvard Annals, 
50, which was the chief source of information regarding stellar spectra 
before the appearance of the Henry Draper Catalogue. Each plotted 
point in figure 1 shows the position of a star of spectral type B (the 
subdivisions B8 and B9 are not classed with the other B's — a customary 
procedure in statistical discussions of spectra). The full curve repre- 
sents the galactic equator, that is, the central Hne of the Milky V/ay 
after correction for the small dip due to the sun's position slightly to 
the north of the galactic plane. The broken curve shows the projection 
of the plane of the local cluster. The determination of this curve was 
based upon stars brighter than magnitude 5.5 {Mt. Wilson Contr., No. 
157, p. 21); if the stars to the sixth magnitude, as plotted in figure 1, 
had been used in the computation, it is very probable that the curve 
would be sKghtly nearer to the galactic equator. 
It is clear at once from figure 1, that these B-type stars are afiiKated 
with the local cluster. We may reasonably assume that about twenty 
of them, along the galactic plane, may be actual members of the galactic 
field, and that the other three hundred are members of the subordinate 
local system, quite unrelated to the plane of the Milky Way. On the 
