218 
DR. C. CHREE: DISCUSSION OF KEW MAGNETIC DATA 
As fig. 2 shows, the difference between sunspot maximum and minimum is mainly 
a matter of amplitude ; but in winter, in particular, the tendency in the afternoon 
curves to a plateau— -i.e., the maintenance of a uniform value—is especially charac¬ 
teristic of sunspot minimum. Active regular changes in H at midwinter are mainly 
confined to the hours 2 a.m. to 2 p.m. Whether we consider the range or the A.D., 
December and January are clearly the months in which the diurnal inequality is 
least. In the case of the eleven years and the sunspot maximum group of years, the 
minimum is in December, but in the S minimum group it falls in January. The 
differences, however, between the two months are not so decisive as to justify the 
conclusion that what is true of this particular eleven years is invariably true. The 
sunspot minimum group in particular, it should be remembered, contained only three 
years. The maximum amplitude appears in June and July ; but May and August 
are not far behind, especially in the sunspot minimum group of years. In sunspot 
maximum years April closely resembles May as regards range and A.D., but it 
exhibits a decided tendency to the evening plateau and morning minimum charac¬ 
teristic of winter. In sunspot minimum years April falls slightly behind September 
as regards amplitude, while September in all years falls markedly short of August 
and exhibits a distinct trace of the winter characteristics. These features, in 
conjunction with the desirability of having the same number of months in each 
season, seem to justify the division of months adopted. Any grouping which 
combined February with March, or October with November, seems less appropriate. 
One or two features in Table VIII. call for remark. Whether we take the range 
or the A.D., the amplitude of the mean diurnal inequalit}^ for the year was greatest 
in 1893, the year of largest sunspot frequency, and least in 1890 and 1900, the years 
of least sunspot frequency. As will be seen presently, the inequality range and the 
sunspot frequency show a very close parallelism in their variation. The changes in 
the type of the inequality from year to year are small, and their elucidation requires 
a more sensitive method, such as the analysis into Fourier waves presently discussed. 
A tendency can, however, be recognised for the hour of the forenoon minimum to be 
slightly earlier when sunspots are few than when they are numerous. 
§ 8. The V inequality data for the 12 months in Table IX. are shown graphically 
in fig. 3, while fig. 4 contrasts the diurnal inequalities for the seasons and the year 
derived from the whole eleven years and from the sunspot maximum and minimum 
groups of years, as given in Tables IX., X., and XI. The V inequality data are not 
quite so smooth as those for H, especially in winter. The fact that the H curves 
were smoothed, while the V curves were not, presumably partly accounts for this. 
The variation in the type of the inequality throughout the year is less in V than in 
H. In some months two distinct maxima and minima are visible in V, but the 
principal maximum always occurs in the afternoon between 4 p.m. and 7 p.m., ivhile 
the principal minimum, in the forenoon, is as uniform in its time of incidence as the 
corresponding minimum in H. The time intervening between the morning minimum 
