Till-; MOVKMKNTS OF TIIK ST A KS. 



[The (Evidence was shown by means of a number of diagrams 

 summarizing the statistics of the proper motions of the stars, 

 each diagram corresponding to a different region of the sky. 

 The diagrams show thai in each case there are two favoured 

 directions of motion along which the stars tend to move more 

 predominantly than in other directions. For comparison, it 

 was shown what kind of a diagram we should obtain if the stars 

 were moving haphazard (except for the solar motion already 

 mentioned) ; and it was pointed out that this was entirely 

 different from the diagrams representing what is actually 

 observed. Finally, a method of dissection was explained by 

 which we can show that the observed motions would correspond 

 very closely to two streams moving in definite directions.] 



I have shown you diagrams for eight different regions of the 

 sky, and in this particular case the whole sky was divided into 

 seventeen equal regions. I daresay there may be some even in 

 this audience, cynical observers of human nature, who may 

 suspect that I have picked out the best examples to support 

 what I am trying to persuade you, and that the other nine 

 regions might not look quite so convincing. You know the 

 frailties of scientists. Well, you are right. Three of the 

 remaining regions show the effect well enough ; but in the 

 other six there is very little that suggests two streams. But if 

 I hold out my arms to represent two streams, from in front you 

 see them plainly separated ; imagine them looked at from all 

 points, above, below, to one side and the other — as we look at 

 the star-streams in different parts of the sky. There are some 

 points of view from which the two arms would come into line and 

 not be seen separated. The six remaining regions of the sky are 

 just those regions where the point of view necessarily confuses 

 the two star-streams — where they come more or less into line. 

 The fact that we do not separate them plainly in these regions 

 is all in favour of the theory, and confirms us in believing we 

 are on the right track. 



It is interesting to plot on a globe the directions of the two 

 streams, found in the different regions of the sky discussed in 

 these investigations. If we pick out one of the streams — 

 Stream I — the directions found for it all converge to a point on 

 the globe. That means that the Stream I directions are 

 parallel, or nearly so, in all parts of the sky. Similarly the 

 directions of Stream II are found to converge. In other words, 

 the different parts of the sky agree in showing the same two 

 favoured directions of motion. 



The result of a number of discussions of different series of 



