PROBLEMS OF MODERN SCIENCE 



12" +50 c 



Antapc* 



and the elongation increases as the speed of the 

 base point is increased. Mathematical theory 

 enables the shape of these curves to be calculated 

 on the assumption that the stars are moving in a 

 haphazard manner and that there is no preferential 



direction of motion. 



But when diagrams are 

 constructed representing 

 the actual number of stars 

 which are observed to be 

 moving in different direc- 

 tions, the figures which 

 are obtained bear little 

 resemblance to those just 

 described (Fig. 4). 



The forms of the 

 Fig. 4 . -curves showing the curves change as we pass 



rr M oio ST i? Dr T ^ one reg ion f the S k y 



Directions to an ^joining One, but 



all have this feature in common, that instead of 

 being elongated in one direction only, they show 

 elongations in two and sometimes three directions. 

 It is clear then that our simple a priori hypothesis 

 regarding the nature of the stellar motions is not 

 in accordance with the facts, and must be re- 

 placed by some other. Now it has been shown 

 by Eddington and others that the main features 

 of the curves deduced from observation can be 

 reproduced by combining in a suitable manner two 



58 



Antaptx 



