722 PHILOSOPHICAL TRANSACTIONS. [aNNO 17g6. 



superiority which is to be considered as of no use in estimations that are intended 

 for the purpose of detecting changes. 



In a foregoing paragraph we have said that this method of ascertaining the 

 histre of the stars was difficult and laborious. The difficulty consists in avoiding 

 the various causes of error that may bias our judgment in assigning the com- 

 parative brightness of the stars : the different altitudes at which we view them : 

 the state and situation of the moon : the time of the night with regard to twi- 

 light : the uncertainty of flying clouds: the twinkling and continual change of 

 star-light, to whatever cause it may be owing ; I mean such changes as last but 

 few moments, or at most but a few minutes: a return into the dark after having 

 been writing by candle-light: the zodiacal light: aurora borealis : and dew or 

 damp on the glasses or specula when a telescope is used. All these, it must be 

 confessed, are real difficulties, which it requires much attention and perseverance 

 to get the better of. Tiiat the method is also laborious may be easily conceived: 

 for each star must at least have two other stars to be compared with, and even 

 these will often be found not to be sufficient. To look out for such proper ob- 

 jects, and then to make the necessary comparisons for every star in the heavens, 

 can be no easy task, especially when we remember the difficulties before enume- 

 rated, to which every single estimation of comparative brightness is subject. This 

 however ought not to discourage us from a work which has in view the investiga- 

 tion of a point of great importance ; and as I have already made a considerable 

 progress, I shall give the result of my labour in small catalogues. 



That these investigations are of the importance we have ascribed to them, will 

 appear when we call to our remembrance the great number of alterations of stars 

 that we are certain have happened within the last two centuries, and the much 

 greater number that we have reason to suspect to have taken place. If we con- 

 sider how little attention has formerly been given to this subject, and that most 

 of the observations we have are of a very late date, it will perhaps not appear ex- 

 traordinary were we to admit the number of alterations that have probably hap- 

 pened to different stars to be 100 ; this compared with the number of stars that 

 have been examined, with a view to ascertain their changes, which we can hardly 

 rate at 3000, will give us a proportion of I to 30. But we are very certain that 

 had a number of observers applied themselves to the same subject, which is of 

 such a nature as to require the attentive scrutiny of many diligent persons at the 

 same time, many more discoveries might probably have been made of changeable 

 and periodical stars, wh.ose variations are too small to strike a general observer. 

 In the application we shall make of this subject however, a proportion, such as 1 

 to 30, or even 1 to 300, is sufficiently striking to draw our attention. 



By observations, such as this paper has been calculated to promote and facili- 

 tate, we are enabled to resolve a problem not only of great consequence, but ill 



