PRESIDENTIAL ADDRESS SECTION A. 41 



us to measure the system as a double star, and so determine the 

 inclination. From these combined observations of real dimensions 

 and apparent size we can find the distance of the pair. The period 

 of revolution varies enormously, from a few years to hundreds of 

 years in the case of visual binaries, and from a few years to a few 

 days, or even a fraction of a day, in the case of spectroscopic 

 binaries. 



The actual dimensions of the systems also van 7 enormously. 

 In the case of spectroscopic binaries of very short period of revolu- 

 tion it is difficult to believe that two separate stars are involved at 

 all, and it is thought that pulsations or alternate expansions and 

 contractions in a single diffuse gaseous star may give rise to the 

 observed oscillations of the lines of the spectrum. From this pul- 

 sating stage we may pass to pairs of stars revolving in contact and 

 then to pairs becoming wider and wider apart as they attain greater 

 age and later types of spectrum. Aitken concludes that "at least 

 one in every eighteen, on the average, of the stars in the northern 

 half of the sky which are as bright as the ninth magnitude is a 

 close double star visible with the Lick thirty-six inch refractor,'' 

 whilst Campbell states that at least one star in three, on the 

 average, of very early type stars is a spectroscopic binary, and of 

 the middle and later types one star in six is a binary. There is 

 still another method of determining a star's distance. It depends 

 on the possibility of measuring the position angle and distance 

 apart of the components of a double star (of known spectroscopic 

 orbit), which is too remote to be perceived as double in our most 

 powerful telescopes. Michaelson has devised a method of making 

 these measurements by the use of an interferometer, and Anderson 

 has applied the method to Capella. 



The value of Michaelson's method lies in the fact that we can 

 extend our measurements much further into space, and deal with 

 stars which could not be reached by ordinary micrometrical 

 methods. Experiments in this direction are now being made at 

 Mount Wilson with the 100-inch telescope. In the light of the 

 foregoing information we can now compile a short list of stars to 

 illustrate the fact that the stars differ enormously in absolute 

 magnitude, the range in intensity being at least from 100,000,000 

 to 1, our own sun occupying a middle place between the brightest 

 star (10,000 times brighter) and the faintest star 10,000 times 

 fainter (see Table II.). Dr. Shapley concludes that if we could 

 place our sun amongst the stars of the Hercules cluster (Messier 

 No. 13) it would appear fainter than the 22nd magnitude, which 

 means that it would be absolutely invisible in the 100-inch reflector 

 at Mount Wilson, the most powerful telescope in the world, and 

 it would be in company with thousands upon thousands of stars 

 of so much greater brilliancy that they could easily be seen in a 

 small telescope. No star as faint as our sun has yet been photo- 

 graphed in this cluster, and all the stars in it which he has cata- 

 logued are more than 200 times as bright. Our brightest star 

 Sirius, apparent magnitude —1-6, placed at the distance of M13 

 would appear fainter than the 18th magnitude. 



