720 PHILOSOPHICAL TRANSACTIONS. [aNNO )7Q6. 



August id, 1783, being on the same subject, of assigning comparative mag- 

 nitudes, I introduced lines to show the intended distances of the letters, with a 

 view to prevent mistakes that might be made in transcribing them, and expressed 



the order as follows: " Order of the stars in Auriga; a ^(3 iS int^ 



UTTT." The marks denoted that all the stars were in succession, but that 



the distance between those which are separated by lines was greater than that be- 

 tween the rest. When stars occurred that were nearly equal, I placed them under 



each other, thus: " Order of the stars in Ursa Minor, a|3 y i 



But in this expression there is the inconvenience of its breaking in on the 

 lines above and below. Another cause of disorder arose from the stars which are 

 not lettered. For here we are obliged to use numbers instead of them; and 

 these, unless properly separated, will run into each other, and occasion mistakes. 

 In the next place, the letters themselves became troublesome; for a star cannot 

 , be found so readily in a catalogue or in an atlas by a letter, as it may be by a 

 number. 



The inconveniences attending the above different ways of notation having now 

 been sufficiently pointed out, it remains only to lay down the method on which, 

 after many trials, I have fixed, in order to avoid them. Laying aside the letters 

 entirely, I use only numbers in all my observations, and these numbers are such 

 as I have added with red ink both to the edition of 17'25 of the British catalogue, 

 and to the Atlas Coelestis taken from that catalogue, and printed in 1729. When 

 I use other stars than what are contained in the British catalogue, the authors 

 who have given them, and their numbers in the catalogues from whence they 

 are taken, are particularly mentioned. 



In the choice of the stars which are to express the lustre of any particular one, 

 my first view is directed to a perfect equality. When two stars are perfectly 

 alike in brightness, so that by looking often and a long while at them, I either 

 cannot tell which is the brightest, or occasionally think one the larger, and 

 sometimes, not long after, give the preference to the other, I set down their 

 numbers together, only separated by a point. For instance, 30.24 Leonis. 

 However, it can happen but very scUlom that the equality in the lustre of two 

 neighbouring stars is so perfect as not to leave an inclination to prefer one to the 

 other; I therefore place that tirst which may probably be the larger, even 

 thougli I do not particularly judge it to be so. But this preference is never to 

 be understood to extenti so far as to make it improper to change tiie order of the 

 two stars; and the expression 24 . 30 Leonis will be equally good with the former. 

 When a third star is concerned, such as 30 . 24 . 77 Leonis, the order of them 

 ought not to be changed; notwithstanding an equality between each member of 



