DISCOVERY 



321 



and the most distant the faintest, and the original 

 assumption of the earher astronomers mentioned before 

 would have been correct. But stars differ very con- 

 siderably in intrinsic brightness. If, therefore, all the 

 stars were placed at the same distance from the Earth, 

 they would appear of different magnitudes, or would 

 have, as it is called, different " absolute magnitudes." 

 The unit of distance adopted by astronomers is that 

 distance where the parallax would be o"-i, and the 

 absolute magnitude of any star is its apparent magni- 

 tude when reduced to this unit distance. 



Now if the distance of a star is known, as it may be 

 by one of the methods previously described, and if, 

 also, its apparent magnitude is known, as is the case 

 with most stars, then it is quite easy to determine the 

 star's absolute magnitude. 



How it is Employed 



The principle on which the new method is founded 

 is that the intrinsic brightness of a star has an appreci- 

 able effect on its spectrum. Thus, if two stars have the 

 same type of spectrum or have approximately the same 

 temperature, but differ greatly in luminosity, they will 

 probably differ greatly in size, density*, and in depth 

 of their surrounding gaseous atmospheres. If this be 

 so, then their spectra should exhibit variations in the 

 intensity and character of such lines as are peculiarly 

 sensitive to the physical conditions of the gases in 

 which they find their origin, in spite of the general 

 correspondence between the two spectra. 



It has been found by Hertzsprung and Adams and 

 Kohlschiitter that certain lines in steUar spectra do 

 give indications of variation with absolute magnitude. 



An examination of these spectra shows that certain 

 lines are more intense if a star is of great absolute mag- 

 nitude than if it is of small absolute magnitude. By 

 measuring the intensities of these lines with respect to 

 other lines in the spectrum, it is possible to determine 

 the true absolute magnitudes of the stars. Having 

 obtained thus the absolute magnitude, and knowing 

 the apparent magnitude, then the parallax can be 

 easily determined. 



The first step in the process is to have available 

 a classification of star spectra like that previously 

 mentioned based on detailed measurements of line 

 intensities. Such a classification already exists and 

 is rapidly being extended. It is next necessary to 

 construct a series of reduction curves for each type or 

 class of spectrum, or for small groups of types. These 

 curves are based on the calculation of absolute magni- 

 tudes of stars from known data such as the apparent 

 magnitudes, and the trigonometrical parallaxes and 

 the relative intensities of selected pairs of lines care- 

 fully measured. 



Having now formed these curves, it is then a simple 



matter to measure the intensities of special lines in 

 any star spectrum and determine the star distance. 



Naturally greater accuracy is obtained when more 

 than one photograph is examined and several pairs 

 of lines in them are used, but this involves very little 

 extra labour. 



The rapidity with which the determinations of 

 parallax can be secured, when once the fundamental 

 curves are formed, is far in excess of that of the older 

 methods. 



The large powerful telescopes of the present day are 

 capable of photographing the spectra of very faint 

 stars, so that a rapid survey of the whole heavens, at 

 any rate to stars of about the seventh magnitude, will 

 be accomplished in the near future. 



Any observatorj' that possesses a large number of 

 stellar spectra can utilise them for this purpose. 



A little more than a year ago the spectroscopic 

 determinations of steUar parallaxes were confined 

 entirely to the United States, at the Observatories of 

 Mount Wilson and Harvard College. The Astro- 

 physical Observatory at Victoria, B.C., has recently 

 proposed to work up their spectra in this research. 



In this country the only observatory that has been 

 and still is occupied with this work is the Norman 

 Lockyer Observatory at Sidmouth. but the Director 

 of the Stonyhurst College Observatory has recently 

 announced his intention of taking up this subject. 



At the Sidmouth Observatory the parallaxes of 

 500 stars have already been determined and published 

 by the Royal Astronomical Society, and another 

 similar number will soon be completed. 



An interesting point in connection with this observa- 

 tory's work is that the measurements of the intensity 

 differences between pairs of lines are being made by a 

 method, originated by the writer, which is different 

 from either of those used at the American observa- 

 tories. Thus a comparison between the American 

 and Sidmouth values is of special interest. 



This opportunity may be taken to point out the valu- 

 able assistance which was given by the Department of 

 Scientific and Industrial Research which rendered possi- 

 ble the undertaking of this research on so large a scale. 



SCIENCE, PHILOSOPHY, AND RELIGION 



In an age such as ours, when men are striving to extract some 

 meaning out of the confusing experiences and impressions of 

 life, the need for books explaining in a popular yet authorita- 

 tive way the relative values of science, philosophy, and religion, 

 and the manner in which these three separate branches of 

 human knowledge and mental and emotional acti\ities can 

 be brought to bear on the past, present, and future problems 

 affecting mankind, is very great. We think that this need is 

 well supplied by a new Library of Philosophy and Religion, 

 edited by Dr. W. Tudor Jones, the first three volumes of which 

 have recently appeared — Metaphysics of Life and Death, by 

 Dr. W. Tudor Jones ; Buddhism and Christianity, by Dr. J. 

 Esthn Carpenter ; and A spects of the Study of Society, by 

 Mr. R. T. Evans. (Hodder & Stoughton, Ltd., 3s. 6d.) 



