38 PRESIDENTIAL ADDRESS SECTION A. 



drift" of the stars. It is evident that the nearer the star, ihe 

 greater will be the "drift" just as in travelling by rail through a 

 forest, the nearest trees appear to move fastest and the distant 

 trees slowly. The measurement of a star's distance by its apparent 

 cross motion is, however, complicated by the fact that all the stars 

 are moving and the disentangling of the parallactic drift and the 

 peculiar motions of the stars is a matter of some difficulty. 



Another important advance has been made by Professor 

 Adams, of Mount Wilson Observatory. He has found by a careful 

 examination of the spectra of stars of known distance — and there- 

 fore known absolute magnitude — that if these stars are arranged 

 in order of absolute magnitude then the spectra also fall into a 

 regular sequence. It is, therefore, only necessary to examine the 

 spectrum of a star and find its position in this sequence to find its 

 absolute magnitude, and by comparing this with its apparent 

 magnitude the distance of the star is found. 



Prof. Adams and his associates have just published (January, 

 1921) a list of the parallaxes of 1,646 stars derived by this spectro- 

 graphs method. It is important to note that the actual distance 

 of the star is no obstacle in this method as no angular displacement 

 has to be measured. This list contains stars with parallaxes as 

 low as 0001" equivalent to a distance of 3,260 light years, and 

 yet these stars, being bright enough to yield a spectrum, must be 

 counted as among the nearer stars. This method fails when a star 

 is too faint to yield a spectrum in our most powerful instruments, 

 but we are not without other methods of ascertaining something 

 about a star's spectrum even when the spectrograph fails. We 

 have now to fall back on the colours of stars as a guide to their 

 actual spectra. The eye and the ordinary photographic plate are 

 most sensitive to different regions of the spectrum. The maximum 

 sensitiveness of the eye is to rays in the yellow and green region, 

 whilst the photographic plate "sees" best in the blue and violet. 

 Stars rich in blue and violet rays therefore appear brighter in star 

 photographs, and red stars appear fainter than we actually see 

 them. The sequence of star spectra corresponds to the sequence 

 of colour, and we have a regular gradation from the bluest stars 

 to the reddest stars. 



By the use of suitable colour-sensitive plates and absorbent 

 screens we can obtain photographs of stars much as we see them 

 as regards relative magnitude. The magnitudes thus obtained are 

 called "photo-visual" magnitudes, and by comparing these with 

 the photographic magnitudes we obtain an index — the "colour 

 index" — to the star's spectrum, and therefore its absolute magni- 

 tude, and by comparison with its apparent magnitude, its distance 



Another very important method of finding a star's distance 

 has recently been developed by Dr. Harlow Shapley, also at Mount 

 Wilson, which can be applied to stars so faint that the spectro- 

 graphic method fails. 



Certain short-period fluctuations in the brightness of giant 

 stars of high luminosity have been carefully investigated and found 

 to be peculiar to these stars. A relation between the period of the 



