270 



CARNEGIE INSTITUTION OF WASHINGTON. 



smaller, on the average, than those given by Kapteyn and by Campbell 

 for stars of all proper motions together. On the other hand, it had 

 already been found that among the stars with large proper motions there 

 occurs an excessive number of great radial velocities. It therefore 

 becomes desirable to find the law which connects the average radial 

 velocity with the corresponding average astronomical motion. The 

 same investigation can easily be made to furnish, besides, a more con- 

 clusive test of the two star-stream theory than has yet been given by 

 the radial velocities. 



The investigation was extended not only to the stars observed at 

 Mount Wilson, but also to the F, G, K, and M stars of Campbell's 

 Catalogue. The following summary for the G stars will serve as an 

 example of the first results of the investigation : 



Mean radial velocities of G stars in kilometers per second. 



The radial velocities have been freed from the sun's motion, and 

 were all taken positively; X is the distance to the nearest true vertex 

 (18** 12'", -12°; or 6'^ 12", + 12°). The numbers in parentheses show 

 the number of stars on which each result depends. For the elements 

 of the sun's motion, apex = 18'' 0"", + 31°; velocity = 20.0 km. were 

 adopted. The existence of the two star-streams is well shown by the 

 increase in mean radial velocity with diminishing X. The proportion 

 of the numbers in the third and fifth columns is about what we should 

 expect according to considerations based on the proper motions. The 

 fact that this proportion also holds for the stars of very small proper 

 motion is very important. It shows, even more strongly than earlier 

 considerations, that the phenomenon of the two star-streams is not 

 confined to the neighborhood of the sun, as has been suggested, but 

 extends to the most distant regions for which we have the necessary 

 data. The other spectral classes show perfectly similar results. There 

 are one or two irregularities in the case of the K stars, the cause of 

 which has still to be elucidated. 



The gradual increase in the numbers of the last column is no less 

 evident than the fact just considered. The phenomenon has not j^et 

 been exhaustively investigated; but this much may now be confidently 

 stated: at least part of the increase, and perhaps the whole of it, is 



