42 



Walter Gyllenberg 



TABLE XIU. 



The mean velocities in the three main directions in 

 the galactic system. 



Spectral 



iber I 

 tars I 













y 



Class 



G M 

 D , 





20(1 





020 





^002 





O 















B 



247 



1.863 



+ 0.549 



2.002 



+ 0.216 



1.268 



(+ 0.162) 



A 



263 



3.932 



+ 0.260 



2.338 



+ 0.754 



2.676 



+ 0.519 



F 



237 



4.54(1 



+ 0.338 



3.376 



+ 0.299 



3.306 



+ 0.696 



G 



208 



4.921 



+ 0.387 



2.643 



+ 1.063 



4.549 



+ 0.877 



K 



486 



4.480 



+ 0.191 



3.979 



+ 0.367 



4.149 



+ 0.332 



M 



85 



4.362 



+ 1.029 



4.829 



± 2.641 



4.437 



+ 2.013 



iiiagn < 4.9 



1069 



4.178 



± 0.093 



2.631 



± 0.244 



3.565 



± 0.186 



39. With the exception of the class B stars, the mean velocities in the three 

 directions agree very well with the results found from the solution of the velocity 

 ellipsoid, where the spectral classes were taken in groups. Concerning the Class B 

 stars, their velocity distribution seems in opposition to the other stars to be repre- 

 sented by an ellipsoid, flattened in the plane of the Milky Way. I will however 

 discuss this fact in another place. 



The relative magnitude of the mean velocities in the separate spectral classes 

 are the same as before, a fact that is very striking. It seems as if the mean velocity 

 in the direction of the principal vertex, except for the B class stars, were almost 

 constant and the mean velocities in the other directions gradually increase when 

 passing through the spectral classes from A to M. In the M class the three axes are 

 equal, the velocity distribution is here almost spherical, quite in agreement with the 

 proposition pronounced by Charlier that the velocity distribution of stars with in- 

 creasing age must necessarily accept a globular shape. 



But how can this circumstance be connected with the theory of giant and 

 dwarf stars, and the existence of two types of age in each spectral class? Especially 

 concerning the class M stars there exists a contradiction. These stars necessarily con- 

 tain a majority of giant stars, at least when, as here, only the bright stars are consi- 

 dered. Is this really the case the globular velocity surface of the class M stars 

 will represent the distribution of the velocities of the youngest of the stars. 



A peculiarity that, as mentioned, seems quite striking, is the fact that the ve- 

 locity ellipsoids have their shortest and longest axes situated in the plane of the 

 Milky Way. The velocity ellipsoids are flattened perpendicular to this plane and not 

 flattened along it, as was found from proper motion observations. It is however 

 not difficult to overcome this contradiction as far as the proper motion results are 

 concerned. A variation in terms of galactic latitude of the mean parallaxes of the 



