86 
DE. C. CHREE: SOME PHENOMENA OF SUNSPOTS AND OF 
while in the latter case it would be 
(198-Gl)x 1786/100 = 2447. 
Thus, on the above hypothesis, the difference between the greatest and least 
representative daily magnetic ranges——out from the days included in Table IV. — 
should be about 5'8 times as large in the sunspot maximum as in the sunspot minimum 
group of years. 
The rise to the crest of the pulse in Table IV. and the subsequent fall practically 
take place between days n-7 and n + 7, i.e., the pulse extends over about 15 days. 
There would obviously be advantages in making the pulse narrower. This could be 
done, to a certain extent, by reducing the monthly number of selected days. But 
even if we took only the one day of largest spot area of the month, the pulse would 
remain far from wall sided, because the days adjacent to that of largest spot area 
have almost always themselves areas much above the average of the month. On the 
other hand, if one took only 1 or 2 days a month, the number of years would probably 
have to be largely increased to get as smooth a progression as that shown in the first 
line of Table IV. 
The last line of Table IV., dealing with the representative days of least spot area 
of the years of sunspot maximum, gives of course a trough instead of a crest at day n. 
The symmetry with respect to column n is not quite so good as in the case of the 
representative days of largest spot area from the same group of years, but still is very 
fair. 
§ 11. Table V. gives the results for H ranges corresponding to the sunspot data in 
Table IV. The values of the ranges in columns n—15 to n+15 are expressed as 
percentages of the mean range derived from these 31 columns. These absolute mean 
ranges appear in the last column of the table. 
If there were no relation between sunspots and magnetic phenomena on individual 
days, then we should expect the departures from 100 in the figures in the first 31 
columns of Table V. to be small and irregular. There is certainly less smoothness in 
the progression of the figures than in Table IV., and the percentage variations shown 
are much smaller; but there are features which it is impossible to ascribe to accident. 
Let us first consider the data from the 11 years in the first four lines of Table V. 
Accidental features are obviously not entirely eliminated, but there cannot be two 
opinions as to the existence of a marked pulse, the crest occurring some days after 
that of sunspot areas. In every case there is a marked depression or trough some 
days in advance of the day n which represents the maximum of sunspot areas. In 
the first line of Table V., representing all months of the year—represented 
graphically in the broken-line curve of fig. 1—the trough occurs about 4 days in 
advance, and the range is still helow its mean on the day before sunspot maximum, 
i.e., at a time when, as appears from Table IV., sunspot area is almost double its 
mean. The rise from the trough to the crest is regular, and the crest itself 
