ON STAR STREAMING. 259 
If, however, our hypothesis of random orientation is approximately 
true, the divergences will be small. The sum of the proper motions at 
right angles to these circles will be nearly zero in every part of the 
sky. 
Not only that, but we will have further to expect that any other 
eondition of symmetry will be approximately fulfilled, too, for every point 
oer 
OS Pole Galaxy 
Arrows point to the apparent vertices : 
A= 85° D= —-11° 
260° —48° 
obtained from direct and retrograde p.m. respectively. 
of the sky. Such another condition will be, for instance, that on both 
sides of the great circles through the apex the total quantity of proper 
motions shall be the same, or, again, that Sa shall be the same on both 
sides of these circles. 
How the first of these conditions is satisfied ig shown in fig. 2. This 
figure summarises the more important points in regard to the question 
82 
