MAGNETIC DISTUEBANCES AT KEW. 289 



in December (representing winter) to 1'58 in March and in June (respectively repre- 

 senting equinox and summer). But on disturbed days we see from the figures in 

 Table IX. that this ratio has risen to 2'30 iu winter and fallen to 0'80 in summer. 



The seasonal variation in the shape of vector diagrams the curves illustrating the 

 diurnal changes in the horizontal components D and H or W and N is thus totally 

 different for quiet and disturbed days. 



12. Taking the year as a whole, the ratio borne by the diurnal inequality range 

 in W to that in N is only 1'18 on disturbed days as compared to 1'46 on quiet days. 

 This is due to the fact that the increase in the range of the mean diurnal inequality 

 as we pass from quiet to disturbed days is 91 per cent, in N as against 54 per cent, 

 in W. In this instance the change in passing from quiet to disturbed days is in the 

 same direction as that observed on quiet days when we pass from years of few to 

 years of many sunspots. 



The mean diurnal inequality for the year in N is, on the whole, of the same type 

 on disturbed and quiet days. The hour of principal maximum, 7 p.m., and principal 

 minimum, 11 a.m., are the same in the two cases. 



There is much more difference of type in the case of W. The principal minimum 

 appears near midnight on disturbed days as compared with 8 a.m. on quiet days. 

 The most conspicuous turning point, however, the maximum, occurs in both cases in 

 the early afternoon, though somewhat later on the disturbed than on the quiet days. 

 The nature of the differences between quiet and disturbed days can best be followed 

 by reference to fig. 2, p. 290. 



13. The increase in diurnal inequality range on disturbed days in D and H, though 

 considerable, is quite overshadowed by that in V. Comparing individual months in 

 Table VII. with the corresponding monthly data on quiet days ((A), Table XIV., 

 p. 363), we find that the range from disturbed days bears to that from quiet days a 

 ratio whose least value is 2 '20 in October, and whose mean values are 2 "91 for 

 equinoctial, 4 '05 for summer, and 6 '31 for winter months. The mean of the ratios 

 from the twelve months, 4 '42, is more than double the corresponding means for D 

 and H. The size of the disturbance influence appeals to the eye in the V curves in 

 fig. 2. These show the diurnal inequalities for the whole year and for winter and 

 summer for both disturbed and quiet days, as well as an inequality derived by 

 subtracting quiet day hourly values from disturbed day values. This last inequality 

 may be regarded from one point of view as representing the disturbance element pure 

 and simple. 



It will be observed in Table VIII. that the excess in the inequality range, shown 

 by the smaller number of more disturbed days, is much more conspicuous in V than 

 in H. The suuspot maximum years show an even more enhanced inequality range in 

 V, especially in winter. 



14. The type of the diurnal inequality on disturbed days in V, as shown in 

 Tables VII. and VIII. and in figs. 1 and 2, does not vary much with the season, but 



VOL. cox. A. 2 P 



