146 Notices respecting New Books. 



reasonable degree of exactness, to the actual state of the heavens, so 

 as to serve as a record to future ages sufficiently correct to ensure the 

 detection of changes " such as there is abundant reason to believe a 

 great many stars have undergone within the period of astronomical 

 history." The subject was long ago urged on the attention of astro- 

 nomers by Sir William Herschel, in his papers on the comparative 

 brightness of the fixed stars published in the Philosophical Transac- 

 tions ; and if anything could add force to his arguments, it would be 

 the extraordinary phenomena presented by the star jj Argils already 

 mentioned, and of variable stars in general. 



Sir John states, that after having made trial of various instru- 

 mental contrivances, he is disposed to rely mainly, for the formation 

 of a real scale of magnitudes, on the unassisted judgement of the 

 naked eye. The method which he adopted, and calls the method of 

 sequences, is an extension of Sir William Herschel's method of naked 

 eye comparisons, so modified as to afford a means of educing from 

 it a numerical scale of values of the magnitudes compared. Sir 

 William Herschel's object in forming his catalogues of comparative 

 brightness, was to identify the lustre of each star so as to enable a 

 future observer to satisfy himself whether any change had taken 

 place ; and to this end each star was compared with one or at most 

 two others judged to be equal to it (or very nearly so), so as to 

 establish a system of binary or ternary sequences or equalities. 

 Supposing this to be done for every visible star, it is clear that no 

 change could take place without being detected. For the purpose 

 proposed by Sir John, namely, the formation of a numerical scale of 

 magnitudes, it was necessary to establish more extended sequences, 

 in each of which a considerable interval of the scale should be in- 

 cluded ; and when many such sequences were accumulated, to de- 

 vise a means of combining them into general sequences, including 

 all the stars observed. His method was, generally, as follows : — 



" Choosing perfectly clear nights (which for this purpose are quite 

 indispensable), a succession of stars was picked out by actual inspec- 

 tion of the heavens from the largest above the horizon down to some 

 of considerably inferior magnitude, and noted down in a list, in a 

 vertical column — leaving blank intervals more or less considerable, 

 according as the steps of the skeleton scale so picked out were wider 

 or closer — but taking care that between the skeleton stars arranged 

 seriatim, there should always be an unequivocal descending step of 

 apparent lustre. The business of the night, then, was to fill in as 

 far as practicable the steps of this scale into an unbroken chain of 

 downward gradation, placing each newly-added star by actual 

 judgement and comparison with its immediate neighbours, in its 

 proper order, until the scale became so gradual in its declension that 

 it was no longer possible to insert fresh stars with certainty between 

 its members, in which case they were set down as equal to some of 

 those already noted down." — P. 306. 



The charts referred to during the observations were those of Bode's 

 Atlas, or working copies made from them. In consequence the 

 nomenclature was Bode's, which, faulty as it is, it was then necessary 



