64 PROF. BALFOUR STEWART AND MR. W. DODGSON ON 



Period in years. Mean departure. 



"f 2*0556 



12 2*1490 



"i •• 1*9538 



The values of the iof- and the 1 2-yearly inequalities 

 freed by this means from the influence of each other are 

 found in Table IV. (p. 66 seqq.). 



We next eliminated the influence of these two inequalities 

 from the whole series of quarterly observations, and with 

 the values so corrected obtained the indications given in 

 Table III. of an inequality at 16J years. The value of 

 this inequality is found in Table IV. Besides these three 

 inequalities, there appear to be traces of an inequality cor- 

 responding to 6 years. This, however, will be incorpo- 

 rated in that whose period is 12 years, and ought not, 

 therefore, to be separately considered. 



1 1 . Having thus determined three inequalities with 

 periods of iof, 12, and 16 J years, it becomes a point of 

 interest to know whether we can reproduce the observed 

 results reasonably well by their means. The united effect 

 of these three inequalities is given in Table IV. col. 5. 

 The result is that the dates of observed maxima and minima, 

 as well as observed ranges, are well reproduced by adding 

 together these three inequalities. Nevertheless there is 

 the following peculiarity remaining behind. Let there be 

 two curved lines, the upper line being drawn through the 

 various observed maximum points, and the lower through 

 the various observed minimum points of declination-range. 

 The spaces between these two lines may be taken to denote 

 the observed capacities fofr declination-range corresponding 

 to the various years. 



Again, in like manner, let there be two other curved 

 lines, the upper line being drawn through the various cal- 

 culated maximum points and the lower through the various 



