TRANSACTIONS OF THE SECTIONS. 



43 



diflerences of the two. These latter show, very clearly, the existence of a cycle. 

 The maximum occurs in the year 1844, and the minimum in the year 1848. 



In the Supplement to the Makerstoim obsei-vations, published last year, Mr. 

 Balfour Stewart remarks that the mean yearly values of the magnetic declination 

 " exhibit some indications of a period in their Value." The observations appear to 

 show that the change of the declination from year to year, at Malierstoim, has 

 increased from 1841 to 1855. When this change, and its variation (supposed td 

 increase proportionally to the time) are computed and deducted fr-oni the observed 

 results, the residual quantities clearly indicate the existence of a period, the maxi- 

 vium occun-ing in the year 1846, and the minimum in 1851. 



The Dublin results exhibit evidences of a similar cycle. The probable value of 

 the magnetic declination at Dublin, in any year, is given by the formula 



■f=26°29'^25-5''94xH, 



» denoting the mmiber of the' year reckoned fi-om 1850. And when the probable 

 values for the several years of observation, computed by this formula, are deducted 

 from those actually observed, the diflerences fiumish, like the Makerstoun results, 

 unmistakeable evidences of a period. But it is remarkable that the epochs of the 

 maximum and minimmn occm- about a year earlier than at Makerstoun, the maxi-^ 

 mum taking place at Dublin in the year 1847, and the minimimi at the end of 

 1842, or beginning of 1843. The totail range of this periodical change at Dublui 

 ==2'^24 



It is impossible to avoid connecting these variations with the periods of mag- 

 netic disturbance. The difierence of the observed results for any magnetic element, 

 and the monthly mean corresponding to the same hour, being regarded as the effect 

 pf the distm-bing cause, the mean of these differences, for any period, will serve as-' 

 a measm-e of the 7nean disturbance. The folio-wing Table contams the mean yearly-' 

 values of these quantities at Dublin, in the case of the magnetic declination, for 

 the vears 1840 to 1850 inclusive. The second column contains the diflerences 

 betw^een these yearly means and the means of all ; they show very plainly the 

 existence of the period whose laws have been so frilly traced by General Sabine; 

 The disturbance is gi-eatly less than the mean in the years 1843, 1844, 1845, and^ 



greatly in excess in 1847, 1848. 



^ . \ 



Year. Mean disturbance. Diff. 



