24 Report S.A.A. Advancement of Science. 



adopted by observers in this field of researcli. Indeed it would be 

 utterly impossible to do so, for each observer has his own particular 

 methods, formed and refined by his own individuality ; he has his own 

 peculiar instruments, suggested by and adapted to his own particular 

 methods. 8till thei-e is a certain area common to all observers. And 

 in this common ai-ea there stands out first the need, nay, the absolute 

 necessity, for rigorous determinations of brightness. 



How is the brightness of each star in the sky determined, and 

 what measures or standards of comparison are used in order to arrive 

 at numerical values for these determinations I A star varies through 

 a certain range. How do we measure and evaluate numerically the 

 amplitude of that range ? 



Nineteen centuries ago Ptolemy divided the visible stars into six 

 orders of brightness, called magnitudes. The brightest stars he con- 

 sidered to be of the first magnitude. Modern observers place eighteen 

 stars in this highest order of brilliancy. The faiiitest stars visible to 

 the naked eye — now reckoned to be between two and three thousatid 

 — he regarded as sixth magnitude stars. This classification w-as 

 naturally a rough one : still his ^division of stars into magnitudes 

 remains until this day. 



In the seventeenth century, when a need for more accurate 

 determinations of brightness arose, each magnitude was again divided 

 into divisions of ten. It is this numerical nomenclature which finds a 

 place in all recent star catalogues. 



When extreme accuracy is sought for, a further division into 

 hundredths is adopted, although since a tenth of a magnitude is the 

 faintest difference which even an experienced eye can assuredly dis- 

 cern, this third subdivision presupposes instrumental and observational 

 refinement of a very high order. 



The nineteenth century had not proceeded very far on its way when 

 the important question arose how it was possible to determine the 

 absolute magnitude of any star. Hitherto the magnitudes of stars 

 were determined in a very arbitrary manner. A certain star, say 

 Sirius, was reckoned magnitude 1 ; another star, just on the borders 

 of visibility, magnitude 6. Intermediate values 2, 3, 4 and 5 were 

 determined after a " steps and stairs " fashion ; or, to put it more 

 .scientifically, between these two standards, 1 and G, intermediate 

 magnitudes were determined by a method of set[uences. 



Every astronomer striving after accuracy felt that this was any- 

 thing but a reliable method, and it was little marvel that in some of 

 the earlier magnitude catalogues of the nineteenth century the dis- 

 crepancies between various estimates were many and glaring. Besides 

 the difficulty of accurately estimating intermediate values by a method 

 of sequence.s, there was the added uncertainty that the \alues thus 

 determined might not be absolute values. One could not say that 

 so many times as a I'O star was brighter than a 2'0 star, so many 

 times was a 2*0 star brighter than a 3-0 star, and a 3-0 star than a 

 4-0. star. 



What was wanted, in the language of lighthouse )uen, was to 



