3G PEESIDEXTIAL ADDEESS SECTIOX A. 



of the IJniverse. Early attempts to measure the distance 

 of a star, althoiig-h not successful in their actual object, led 

 to far-reaching discoveries. Bradley, in 1725, in atteniptino- 

 to find the distance of the star Gamma Draconis, discovered 

 aberration and nutation. Herschel discovered the existence 

 of binary stars, and, on investigating the apparent changes 

 in the positions of a few stars from tliose given in early star 

 catalogues, discovered the motion of the Sun, and made a 

 fortunate guess as to its direction of motion and velocity. 

 Ilis first determination depended upon the observations of 

 only seven stars, but does not differ very widely from tho 

 results of recent investigations, in which the ap])arent motions 

 of thousands of stars have been discussed. The application 

 of the spectroscope has confirmed the result, and has added 

 a direct determination of the solar velocity. Stars in a certain 

 direction in the sky appeared to be moving towards the Sun ; 

 others in the antipodes of the sky apj^eared to be receding 

 from us with the same velocity. This velocity has been deter- 

 mined to be 195 kilometres per second. It was soon realised 

 that one result of the Sun's motion through space was to 

 afford a new method of determining stellar distances applic- 

 able to distances so great that the ordinary direct method 

 failed. In one year the Sun, and with it the whole of tliQ 

 solar system, moves through space over a distance equal to 

 four times the distance between the Sun and the Earth, or 

 372 million miles. In ten years the distance moved is 

 forty times the Sun's distance, and the distance increases 

 directly with the time. Thus an indefinitely increasing base 

 line is obtained, whereas in the direct method of determining 

 a stellar parallax the base line cannot exceed 18G million 

 miles, or two units. If, then, an exact comparison can be 

 made between the relative positions of the stars at the present 

 time and their relative positions many years ago, there must 

 be certain differences due to the change in the point of view. 

 The nearer stars lying in directions at right angles to the 

 direction of the Sun's motion must be displaced with reference 

 to more distant stars in the same direction, and generally 

 the stars would appear to diverge from the apex of the Sun's 

 way and to converge towards the solar antajjex. If, now, 

 the stars were all at rest, it would be a very simple matter 

 to measure the distances of thousands and thousands of stars 

 by their displacements relatively to more distant stars, and 

 the number of stars whose distances could be measured and 

 the accuracy of the results obtained would increase as the 

 time increased. But analogy suggests that it is hardly likely 

 that our Sun alone is in motion and all the rest of the stars 

 are fixed in space, and investigation has proved that prac- 

 tically all the stars are in motion. If these motions were all 

 at random, it would still be possible to determine the 

 distances of groups of stars, for the random motions taken 

 over the group would neutralise one another, and the group 

 as a whole would be at rest. In this way Kapteijn, amongst 

 others, has statistically determined the distances of various 

 groups of stars. It had been suggested by Kobold, and later 



