92 DR. C. CHREE: SOME PHENOMENA OF SUNSPOTS AND OF 
other a “ 1,” they are considered side by side, before a decision is reached. This will 
explain the general method. There are, of course, not a few days when the giving a 
“ 0 ” or a “ 1,” or a “ 1 ” or a “ 2 ” is very much a toss up. If the choice made one 
day were to be made independently even the next week, there would be no doubt an 
appreciable number of alterations in the figure ascribed, and the longer the interval 
between the two choices the more would this tend to be the case. 
In view of this fact, in the present enquiry, the days of each year were dealt with, 
so far as practicable, in immediate succession. While the standard remained, I think, 
fairly uniform throughout the days of any one year, it not improbably varied sensibly 
as between different years, for the time that elapsed between the consideration of the 
first and last of the years was naturally considerable. 
In assigning the character figures the appearance of the curves was alone considered, 
and the lists of the daily ranges were never consulted, so that the two lines of 
investigation are at least absolutely independent. A distinction to be borne in mind 
is that the character figure is a measure only of disturbance, whereas the daily range 
is usually dependent both on disturbance and on the regular diurnal variation. 
The data resulting from the use of the character figures appear in Table VII., results 
being given for the 11 years combined and for the same groups of years as in Tables IV. 
and V. The results in the 31 columns w—15 to ?i+15 are expressed as percentages of 
their mean; and the absolute values of these means are given in the last column 
of the table. In the last line of Table VII., as in the corresponding lines of Tables IV. 
and V., the column n contains the representative days of smallest spot area. 
§ 16. The progression of the figures in Table VII. is less smooth than in Table V., 
but the conclusions indicated are very similar. In each of the first five lines, where 
the selected days were those of largest spot area, there is a conspicuous trough a few 
days before the representative day n, followed by a considerable rise to a crest, which 
occurs usually on day n + i. This crest markedly overtops the average mean, except 
for the group of years 1891, 1895, and 1896, which exhibits the same peculiarity as in 
Table V. In addition to this pulse there is a second pulse, as in Table V., with its 
crest about 11 days before the day of largest sunspot area. The chief departure from 
the phenomena seen in Table V. is that this earlier pulse has become decidedly more 
prominent. This fact, and the further fact that the earlier, or secondary, pulse is in 
both tables especially conspicuous in the case of the highly disturbed years 1891, 1895, 
and 1896, suggests that the phenomenon is largely a pure disturbance eflect. At the 
same time, the years of sunspot minimum 1890, 1899, and 1900, show a more 
prominent crest at day w—12 than do the years of sunspot maximum, which were 
much more disturbed. 
An explanation of the secondary pulse which may suggest itself is that a reduction 
of sunspot area below the mean for the year is itself a cause of disturbance. If, 
however, this were the true explanation, the last line in Tables V. and VII. should 
exhibit a crest about day 7i + 4 much more prominent than the crests in any of 
the other lines at day n-ll. This it will be seen is far from the case. 
