104 
DK. C. CHREE: SOME PHENO.MENA OF SUNSPOTS AND OF 
evidence in the sunspot maximum group of years, vrhich gave the most prominent 
indication of the above pulse. The prominence of the 27-28-day period in the sunspot 
minimum groups of years, which contained comparatively few magnetic storms, would 
alone suffice to show that the phenomenon, whatever its iiatuie, is not confined to 
outstanding disturbances such as those chronicled by Mr. Maundee,. 
A special feature in the results for 1891, 1895, and 1896 is a faint indication of a 
shorter period of about 14 days. The figures for this group of years in columns n+l^ 
to ^+16 all exceed the figure in any other column between n + b and n + 2.^. This 
seems hardly likely to be pure accident, and it may represent Prof. Schuster’s 13'64 
days’ period. 
§ 25. In view of the interest attaching to the reality of a 27-28-day period, a similar 
investigation to the preceding was carried out for the two years 1894 and 1895, 
employing the H ranges instead of the character figures. 
These two years were selected because, when character figures were employed, the 
one, 1895, showed the 27-28-day period specially clearly, while the other, 1894, 
showed it less clearly than perhaps any other year. The investigation based on H 
ranges took account only of the 5 representive days a month and the 35 following days. 
The results are given in Table XII., accompanied by the corresponding results previously 
obtained from the character figures. To facilitate comparison, both sets of figures are 
expressed as percentages of the arithmetic mean value derived from the 36 columns. 
In the case of 1894 two sets of H range figures are given. The former and the 
character figures given depend on all the 60 selected days of the year. In obtaining 
the second set of H range figures, all ranges were omitted which exceeded 2OO7, and 
a mean was taken from the remaining figures in each column. The object was to see 
the effect of omitting a few of the larger magnetic storms—all, in fact, which gave H 
ranges in excess of the largest range of 1895. 
The data from 1895, though naturally not as smooth as the 3-year and 11-year 
data in Talde XI., show the 27-28-day period quite as clearly. It is as unmistakable 
in the range as in tlie character figures. Again both range and character figures 
afibrd distinct indications of a 13—14-day period, which it will be remembered 
appeared in Table XI. only in the group of years containing 1895. 
In 1894 the character figures and the range figures (i) both show not a peak at day 
u,H- 27 or 7i-l-28, but a high plateau extending from about day n + 2b to day 7H-34. 
The range figures, in fact, would seem to favour a Schmidt, or 30-day period, rather 
tlian a 27-28-day period. The reason of this is clear on examining the range 
(ii) figures. The high values in the range (i) figures in column n-f29 and subsequent 
columns were maiidy due to the fact that two outstanding ranges, respectively 637y 
and 660y, “happened” to fall one or both in these columns. 
§26. The investigation embodied in Tables XI. and XII., though leading to some¬ 
what unexpected results possessed of intrinsic interest, does not afford an immediate 
explanation of the phenomenon for which an explanation was being sought, viz., the 
