ON STELLAR DISTRIBUTION AND MOVEMENTS. 255 



been overcome (the solution is divergent except under very special 

 conditions). I may mention, however, that I made a rough preliminary 

 solution, which avoided the difficulties of divergency. Much to my 

 surprise, the contour lines for this solution showed two points of 

 maxima and generally resembled the figure G (or rather figure 

 modified by assuming two drifts containing unequal instead of equal 

 numbers of stars). It is hard to say how much weight should be 

 attached to this rough result ; my attempts to make a closer and really 

 trustworthy solution have up to now been unsuccessful. 



Mean Radial Velocities} 3 

 (After correcting for the ' Solar Motion.') 



The phenomenon of streaming in two favoured directions has now 

 been detected in the radial motions by Hough and Halm, and by Camp- 

 bell. Radial motions are rather awkward to work with, because, if 

 you wish to compare motions in two particular directions, you cannot 

 use stars in the same part of the sky, but must compare those parts 

 of the sky for which the directions in question are radial; this some- 

 times introduces a complication. The table, however, shows that 

 near the vertex and anti-vertex (towards which the two favoured 

 directions point) the motions are greater than midway between them; 

 this means that motions in the supposed directions of the streams are 

 on the average greater than those transverse to them, and thus confirms 

 the theory. The difference is not so great as would be expected from 

 our other knowledge of the stream motions ; but there seems to be a 

 satisfactory explanation of this. The vertices lie exactly in the plane 

 of the Milky Way, a part of the sky where early type stars are 

 especially abundant. As already mentioned, these have individual 

 motions considerably smaller than the later type stars, and so the 

 average is unduly lowered in the neighbourhood of the vertices. I 

 think this fully explains why we do not find so big a difference as might 

 have been expected. The variation depending on position in the sky 

 has not been separated from the variation depending on direction of 

 motion, and the one partially counterbalances the other. 



The exceptional character of the motions of the Helium or Orion 

 stars, which was first discovered by Frost and Adams 14 has been the 

 subject of several important investigations. Very recently Campbell 13 



13 Campbell, Lick Obs. Bull, No. 196, Table IV. 



14 YerJces Observatory Publications, vol. ii., p. 1U5. 



15 Lick Obs. Bull., No. 195. 



