116 
DR. C. CHREE: SOME PHENOMENA OF SUNSPOTS, ETC. 
Thus the percentage change of H range calculated as in (2) would have been 
(1-07 X 78-6 X 1-35) -r (1 + 1‘07 X 0786) = 62, 
while the percentage change given in Table V. is 40. Thus the ratio of the observed 
to the calculated change is 0'65. 
§ 33. Even mean annual values of the absolute H ranges do not vary with sunspot 
frequency at all as closely as do the diurnal inequality ranges. Still, if we compare 
any one of the 3 years of sunspot maximum, 1892, 1893, 1894 with any one of the 
3 years of minimum, we find a greater difference between the absolute ranges than 
between the inequality ranges. Thus everything points to the conclusion that the 
magnetic wave, with crest 4 days after that of sunspots, represents only a part—in 
general, probably the smaller part—of the sunspot influence. 
In addition, there seems to be an influence which maintains a high average range 
of diurnal variation in years of many sunspots, so that even when sunspot area for 
several days in succession falls to a comparatively low level, the daily range continues 
to exceed that normal to the same month of a year having the same mean sunspot 
area as the specified days. 
Also there is a more distinctively disturbance element whose amplitude does not 
appear to be directly proportional to sunspot area, and which seems largely responsible 
for the 27-28-day period shown in Tables XI. and XIV. 
It must be remembered that Tables V. and VII. cover only a sufficient number of 
days to show a sunspot influence which acts on the earth within 15 days of the 
sunspot phenomenon on the sun. A period longer than this, but shorter than 
2 months, is not contradicted by Table XVIII. 
Again, sunspots may not themselves be the actual sources of the solar influence, 
but only symptoms that something is happening, has already happened, or may only 
be about to happen, which exerts an influence on the earth. 
