40 PRESIDENTIAL ADDRESS— SECTION A. 



metrical measures of its distance, 155 light years (which means that 

 the outburst occurred in 1763 and we only perceived it in 1918), 

 and its apparent maximum magnitude — 1-5. 



Many Novae have been discovered photographically in the 

 great nebula in Andromeda; they have usually attained the 17th 

 magnitude. If we can assume that they are comparable in 

 brilliance to the one in Aquila, we have a difference of apparent 

 magnitude —1-5 to 17-0=18-5, which, from the known laws con- 

 necting apparent magnitude and distance, corresponds to a distance 

 5,000 times as great as the one in Aquila or 775,000 light years. 

 We conclude therefore that the great nebula in Andromeda is in 

 round numbers 800,000 light years distant. 



This is the same figure as that derived for some spiral nebulae 

 by Aitken by similar reasoning from more extensive data con- 

 cerning Novae in spirals and in the Milky "Way. 



It must be understood that in speaking of the distances of the 

 remoter stars, a few million miles or even a few light years are but 

 as the dust of the balance. 



Very many of the stars which to the unaided eye appear single 

 are seen to be double when examined telescopically. 



In verv many cases the two stars are physically connected, as 

 each of them revolves round their common centre of gravity. 

 Alpha Centauri, our nearest stellar neighbour, is a familiar 

 example; it may be seen to revolve even in small telescopes. The 

 period of revolution is a little over 81 years, whilst other pairs 

 take hundreds of years to perform a revolution. Alpha Centauri 

 is so near that the angular distance apart of the two components 

 varies from 22" to less than 2" during a revolution. More distant 

 binary stars appear as closer doubles, and, for a given absolute 

 separation, beyond a certain limiting distance the two stars cannot 

 be separated by the most powerful telescopes. 



Where the telescope fails the spectroscope is still able to 

 prove that a star apparently single is in reality double, or even 

 triple, the revolution of the two stars round their common centre 

 of gravity being shown by the oscillation of the lines of the 

 spectra about their normal positions. 



Knowing the speed, from the measured displacement, and the 

 time taken to make a complete revolution, the minimum dimen- 

 sions of the system can be determined. If the plane of the orbit 

 is at right angles to the line of sight, the spectroscope cannot detect 

 the rotation, as there is no motion of approach or recession, but if 

 the orbit contains the line of sight, the spectroscopic orbital 

 velocities reach their full value. In this case one star eclipses the 

 other at every revolution, and we have a more or less sudden drop 

 in the brightness. Stars showing this phenomenon are known as 

 eclipsing binaries or Algol variables, the star Algol being the type 

 star, showing a drop in brightness every 2-867 days. At inter- 

 mediate inclinations of the orbits the spectroscopic velocities of 

 rotation are only a fraction of the real ones, proportional to the 

 sine of the angle of inclination, and we have no means of knowing 

 the actual size of the orbits unless the binary is near enough for 



