ON THE MAGNETIC DECLINATION. 



11 



Prague observations also confirm the nearly equal deflections (mean) to the west 

 and east. The epochs of the maxima and minima were found from the four roots 

 of the equation = 0.029 cos (6 + 295) + 0.414 cos (20 + 85), which gave 10 

 minutes as the mean time elapsed between the moon's passing the meridian, and 

 the time of maxima deflections. If we take the four phases into account, the lunar 

 action seems to be retarded 10 minutes, which quantity may be termed the lunar- 

 magnetic interval for the Philadelphia station. At Toronto the intervals are not 

 so regular. 



The secondary range exists at Toronto, and is a marked feature in the Prague 

 result. 



The following table contains the observed and computed values and their dif- 

 ferences : 



The formula or curve enables us to divide the observed curve so as to show the 

 diurnal and semi-diurnal part of the observed variations. The decomposition of 

 the curve is made on the diagram where the resulting curve for the diurnal period 

 is given. 



The lunar-diurnal variation seems to be subject to an inequality depending on 

 the solar year, for the investigation of which the preceding results were rearranged 

 in two groups, one containing the hourly values for the summer months (April to 

 September), the other the values for the winter months (October to March). For 

 the summer season we have 11,087 observations, and for the winter 10,557. 



