1919] on Giant Suns 417 



the voice goes through changes which show a reversal after middle 

 life. We were practically confusing the baby stars with the old man 

 stars until Russell called our attention to the fact ; and now it seems 

 quite easy to make the distinction. But there was some hesitation 

 before the new views were accepted at all, chiefly on account of the 

 lack of sufficient measures of distance, which left room for doubt. 

 Recently the evidence has been reinforced in a remarkable way by a 

 totally new and unexpected method for inferring the distances of 

 stars, due to Mr. W. S. Adams, of the Mount Wilson Observatory in 

 California. His discovery is that if we have two stars, one of which 

 is very bright intrinsically and the other faint, but both of the same 

 spectral type, we can find two lines of the spectra which have 

 different relative intensities : let us call them A and B. In the 

 bright star A is more intense than B, in the faint star B will be 

 more intense than A. Now observe that this difference will persist 

 however far we may remove the stars from us. By altering the 

 distances we may make the brighter star appear the fainter, but we 

 can pierce its disguise by noting simply that the line xV in the 

 spectra is the more intense, so that if the star appears faint we see at 

 once that this must be due to its greater distance. In fact we can 

 infer the distance from the relative intensities of the lines A and B, 

 so that Adams has really given us a new method of inferring dis- 

 tances. The new method has the further advantage of requiring far 

 less labour than the old method of parallax ; in fact, when once the 

 spectrum has been photographed the further labour required is quite 

 small, so that by this time Adams has been able to give us the 

 luminosity of hundreds of new stars, and by this overwhelming 

 evidence confirms Russell's results derived from merely a few. In 

 reply to a request he has sent me specially for this lecture the 

 following table of results, and I am sure you will appreciate his kind- 

 ness. [In the lecture the results were represented diagrammatically ; 

 here Adams's actual figares are reproduced on the next page. To 

 see the " step-ladder " hold the table sideways, so that the column 

 of absolute magnitudes is at the bottom. The upright leg is then 

 represented by the two rules across the page. It will be seen that 

 the majority of intrinsically luminous stars are contained between 

 these lines. The sloping leg is easily seen from the lie of the figures. 

 That there are very few faint stars of class M does not mean that 

 there are few in the heavens, but that they are the most difficult to 

 observe from their faintness.] 



In addition to this confirmation by new observations we have 

 had an independent confirmation by the brilliant theoretical work of 

 Professor Eddington, who has attacked the problem of the life of a 

 star mathematically. He supposed a mass of gas first of all to be 

 simply under the action of its own gravity. It will consequently 

 contract, and owing to the contraction will rise in temperature ; but 

 Professor Eddington soon found that this simple hypothesis would 



