88 
DE. C. CHEEE: SOME PHENOMENA OF SUNSPOTS AND OF 
rounded, the ranges in columns 91 + 3 to n + b being almost equal. The crest value is 
12 per cent, above the mean range, and the preceding trough value 6 per cent, below. 
Thus the amplitude of the wave from trough to crest represents 18 per cent, of the 
average H range. 
The fio-ures for the three seasons in Table V. are less regular. The number of 
years would presumably have to be trebled to get results as smooth as in the first 
line of the table. All the seasons, however, show the same general features as the 
year as a whole. In winter and equinox an excess above the mean range appears at 
an earlier day than in the case of the year. In summer, on the other hand, the 
rano-e does not attain its mean value until the sunspot maximum has passed. The 
differences, however, that exist between the seasonal data for sunspots in Table IV. 
may be responsible for some of the differences between the seasonal data in Table V. 
The crest occurring after the day of sunspot maximum is not the only one visible in 
Table V. The results for the year and the three seasons all show a secondary pulse, 
whose crest occurs about 11 days before the day of largest spot area, i.e., about 15 
days before the crest of the principal pulse in the H ranges. 
Coming to the separate groups of years, each, it will be seen, shows the above two 
pulses, the crest of the one about 4 days after, the crest of the other about 11 days 
before, the crest of sunspot areas. The group of years in which the phenomena are 
most prominent and regular is that including the three years of sunspot maximum. 
In this case the difference between the principal crest and the preceding trough 
represents 40 per cent, of the mean daily range, or about 32y. 
The sunspot minimum years give very similar results to the sunspot maximum 
years, I)ut the difference between the principal crest and trough is only 20 per cent, 
of the mean dally range, or about 9y. The group of years 1891, 1895, 1896 exhibits 
special features. This group includes the years in which the earlier Investigation, 
referred to above, found large sunspot areas associated with small D ranges, and 
conversely. The same conclusion would have followed from the present investigation 
if it had been limited to the day of largest spot area and three following days. The 
fundamental phenomenon apparently was an exceptional development of what in the 
other years is a secondary pulse, with a postponement of the trough preceding the 
day of largest spot area until day n—1. As in other years, however, there is a 
marked rise from this trough ; but the trough itself is unusually deep, and the crest 
of what in other years is the principal pulse does not much overtop the average value. 
Ihus in some ways the depression in the H ranges some days in advance of the 
sunspot pulse appears a more persistent feature than the excess in the H ranges some 
days after the crest of that pulse. 
§ 12. The last line in Table V., answering to the selected days of least spot area, 
dealt with in the last line of Table IV., shows a marked depression in the H ranges, 
extending from 2 days before to 8 or 9 days after the day n of least sunspot area. 
The actually lowest figure is on day 9i+l, but from day n to day n + b the value is 
