710 PHILOSOPHICAL TRANSACTIONS. [anNO 17Q6. 



to them, do more than raise a considerable suspicion of a chqinge. But as this 

 subject will occur again hereafter, and as it must be sufficiently apparent that the 

 present method of expressing the brightness of the stars is very defective, we 

 now proceed to propose a different one. 



I place each star, instead of giving its magnitude, into a short series, con- 

 structed on the order of brightness of the nearest proper stars. For instance, 

 to express the lustre of d, I say cde. By this short notation, instead of refer- 

 ing the star d to an imaginary uncertain standard, I refer it to a precise and 

 determined existing one. c is a star that has a greater lustre than d ; and e is 

 another of less brightness than d. Both c and e are neighbouring stars, chosen 

 in such a manner that I may see them at the same time with d, and therefore 

 may be able to compare them properly. The lustre of c is in the same manner 

 ascertained by ecd ; that of b by abc ; and also the brightness of e by def ; 

 and that of p by efg. 



That this is the most natural, as well as the most effectual way to express the 

 brightness of a star, and by that means to detect any change that may happen 

 in its lustre, will appear, when we consider what is requisite to ascertain such a 

 change. We can certainly not wish for a more decisive evidence, than to be 

 assured, by actual inspection, that a certain star is now no longer more or less 

 bright than such other stars to which it has been formerly compared ; provided 

 we are at the same time assured that those other stars remain still in their former 

 unaltered lustre. But if the star d will no longer stand in its former order cde, 

 it must have undergone a change ; and if that order is now to be expressed by 

 CED, the star has lost some part of its lustre ; if on the contrary, it ought 

 now to be denoted by dce, its brightness must have had some addition. Then, 

 if we should doubt the stability of c and e, we have recourse to the orders bcd, 

 and DEF, which express their lustre ; or even to abc, and efg, which continue 

 the series both ways. Now having before us the series bcdef, or if necessary 

 even the more extended one abcdefg, it will be impossible to mistake a change 

 of brightness in d, when every member of the series is found in its proper or- 

 der, except D. 



Here I have used the letters of the alphabet merely to explain my way of 

 fixing the order of brightness of the stars. In the journal or catalogue itself, 

 which gives this order of brightness, each star m.ust bear its own proper name, 

 or number. For instance, the brightness of the star S Leonis may be expressed 

 by (3iJe Leonis, or better by g4 — 68 — 17 Leonis ; these being the numbers 

 which the above three stars bear in the British catalogue of fixed stars. Per- 

 haps it may be thought that the known introduction of letters, added to the 

 magnitude of the stars, seems to be that very method which I now recommend, 

 as different from what has already been used. And certainly if letters had been 



