1904] Magnetograms on " Quiet" Days, 1891 to 1902. 325 



Kew and at Falmouth. Of the differences between these 12 pairs of 

 dates only three were as large as 2 days, and the mean of the 

 differences taken algebraically was only about 2 hours. The agreement 

 in short was well within the limits of probable error. 



This result again applies to the average year of a sun-spot cycle. 

 Comparing Falmouth results from the two periods, 1892 to 1895 and 

 1899 to 1902, it was found that a small but decided difference existed 

 between the dates of occurrence of the maximum in the annual terms 

 in both declination and horizontal force, and in the semi-annual term 

 in declination. The dates proved to be accelerated in the sun-spot 

 maximum period as compared to the sun-spot minimum period. To 

 make certain that the result was not peculiar to Falmouth, a similar 

 comparison was instituted between the Kew data for the two periods 

 1892 to 1895 and 1890, 1899 and 1900. The results were closely 

 similar to those obtained at Falmouth. Thus the average results for 

 the acceleration in the sun-spot maximum as compared to the sun-spot 

 minimum period from the ranges, the 24 differences from the mean, 

 and the 24-hour term in the diurnal inequality, were as follows : 



The phenomenon emphasises the importance of employing contem- 

 poraneous data when comparing two stations. 



Applying Wolf's formula R = a + 6S, associating the range R of a 

 magnetic element with sun-spot frequency S, results are obtained for 

 the variation of b and b/a throughout the year at Falmouth very similar 

 in character to those previously obtained for Kew. On the whole the 

 Falmouth values of b/a are distinctly the higher, i.e., sun-spot influence 

 is greater at Falmouth than at Kew. 



Taking the above formula, but making S represent not merely 

 Wolfer's sun-spot frequency, but in turn the areas of whole sun-spots, 

 umbrae and faculse as observed at Greenwich, Mauritius, and Dehra 

 Dun, and given by the Astronomer Eoyal in the ' Monthly Notices of 

 the Royal Astronomical Society,' values are calculated for a and b in 

 the case when R represents the range of declination or horizontal 

 force in the mean diurnal inequality for the year. A comparison is 

 then instituted between the ranges for individual years of the 12-year 

 period as calculated from the values of a and b thus found, and the 

 Astronomer Royal's mean yearly data on the one hand, and as actually 



