108 



Prof. B. Stewart on the Daily 



[Mar. 22, 



5. The numbers of Table II. require to be further dealt with before 

 they can be made to furnish a curve, bringing out the long-period varia- 

 tion of the declination-range. Let us first take for this purpose, as well 

 as for other objects to be afterwards mentioned, a series of values derived 

 from the numbers of Table II., each representing the mean of 12 conse- 

 cutive values of Table II. These may be termed three-monthly values. 

 Thus, for instance, we have as follows : — 



Table III. — Exhibiting the Method of obtaining Three-monthly Values. 



Date, 1858. 



Monthly Values 

 for Table II. 



Three-Monthly 

 Values. 



Feb. (3) 



1034 



. 9831 



Mar. (0) 



1022 



[ . . 983 





. 983 J 



» (1) 



1025 



. 9801 



» (2) 



1025 



I . . 977 

 . 974 J 



„ (3) 



988 



. 961] 



Apr. (0) 



952 



L . 955 

 . 950 J 



» (1) 



940 





We have thus, in the last column of Table III., a series of three- 

 monthly values corresponding to the beginning and middle points of each 

 month. In the next place, by adding together a certain three of these 

 values, we may obtain nine-monthly values. Thus the three-monthly 

 value for March (0), as above, is 983, while that for June (0) is 885, and 

 for Sept. (0) 986 ; the mean of these (being 951) is the nine-monthly value 

 corresponding to June (0). Nine-monthly values have thus been obtained 

 corresponding to the beginnings of each month ; and finally, by adding 

 these together, two and two, a series of nine-monthly values have been 

 obtained corresponding to the middle points of each month. These are 

 given in Table IV., and a curve exhibiting them is likewise given in 

 Fig. II. attached to this paper. Again, the numbers given by Messrs. De 

 La Eue, Stewart, and Loe wy in their paper on " Solar Physics " (Phil. 

 Trans. 1870, page 111), exhibiting the spotted area of the sun's visible 

 hemisphere for the years for which we have Kew declination results, 

 have been treated in a manner precisely similar to the above ; that is to 

 say, nine-monthly values corresponding to the middle of each month 

 have been obtained. These values are given in Table V., and a curve 

 exhibiting them is likewise given in Pig. I. (p. 105). 



