28 C. V. L. CHARLIER, 
Taking into account all subclasses and all stars brighter than 
the 6" magnitude we obtain as polar coordinates of the antapex the 
values: 
D = — 339.79 
A= 899.16 
As usually, in all Apex determinations, the declination of the 
Apex is rather vaguely determined. This uncertainty is, without doubt, 
connected with the fact that the principal vertex in the motions of the 
stars has very nearly the same right ascension as the Apex, whereas 
the declinations differ by more than 50°, If the stars in considera- 
tion have a surplus of members moving against the vertex we ought 
to get a small declination of the apex, whereas the contrary leads to 
an increase of this value. 
13. Having determined the value of R for each subclass, we 
are able to calculate the coordinates of each individual star of the 
spectral type B in absolute measure. The discussion of these coordi- 
nates will be carried out with all the carefulness demanded by the 
importance of the problem. 
One of the first conclusions arrived at concerned the law of 
distribution of the stars. It was found, indeed, that this law — so 
eagerly sought for these last years in the discussion of the constitution 
of the Milky Way — was, with great approximation, simply the nor- 
mal frequency law of type A for three variables. Of course this holds 
good, for the present, only for the stars of type B, but we have good 
presumption for expecting the same law for stars of other types. 
If the law of distribution is that of type A, we may determine 
the parameters of the frequency function according to known rules. 
I refer in this respect to Meddelanden N:ris 58 and 66. 
14. Coordinates of the centre’. Let z", y", z" denote the right- 
angular coordinates of a star in Siriometers, referred to the usual 
astronomical (equatorial) system of coordinates, then we have accor- 
ding to 8 4 
c 
DER NY EN) CNE RENTE 
* In this and the following paragraphs (to $ 22) I use all stars in Pickering's Cata- 
logue of B-stars (H. A. 56). Compare S 22. 
