260 
DR. C. CHREE: DISCUSSION OF KEW MAGNETIC DATA 
§ 26. Table XL. gives the ratio borne by the mean B/ to the corresponding arithmetic 
mean of the values of It for the 12 months and for the 11 years separately. The 
disturbed days omitted from the V inequalities were the more numerous, so that any 
reduction in It consequent on this exclusion would naturally be greater in V than 
in H. The excess of Pd is, however, invariably much larger for H than for V, 
confirmatory of what has been already said as to the former being in general much 
the more disturbed element. Relatively considered, the excess of R' over R has 
a conspicuous maximum near mid-winter, and a minimum near mid-summer. The 
pre-eminence of the values of R'/R in winter over those in equinoctial months is hardly 
what one would have expected from consideration of disturbances alone. We see that 
the size of the regular diurnal inequality is apt to give in winter a very inadequate 
idea of the average diurnal variation in the field. 
Table XL.—Values of (Absolute Range from All Days)/(Inequality Range 
from Ordinary Days). 
Element. 
January. 
February. 
March. 
April. 
May. 
June. 
1 * 
August. 
September. 
October. 
November. 
December. 
Mean. 
H 
Y 
2-81 
2-65 
3-25 
2-44 
2-40 
1-73 
1-73 
1-31 
1-72 
1-26 
1-52 
1-18 
1-63 
1-25 
1-67 
1-35 
1-90 
1 -51 
1-97 
1-64 
2-41 
2-10 
2-84 
2-56 
2-13 
1-75 
1890. 
1891. 
1892. 
1893. 
1894. 
1895. 
1896. 
1897. 
1898. 
1899. 
1900. 
Mean. 
H 
1-90 
1-99 
2 ■ 22 
1-74 
2-09 
1-90 
2-11 
1-88 
2-16 
1-97 
1-67 
1-97 
Y 
1-37 
1-54 
2-00 
1-49 
1-82 
1-49 
1-55 
1-32 
1-49 
1-38 
1-32 
1-52 
If we compare the values of R'/R in different years, we see that while on the whole 
it was least in the years of fewest sunspots, it was considerably below average in 
1893, the year of sunspot maximum. In the case of H, in fact, it was lower for 1893 
than for any other year except 1900. This means that in 1893 R was specially large, 
and not that R' was small. As Table XXXYIII. shows, the mean value of R' for 
1893 was 15 per cent, above the average of the 11 years, and was exceeded only in 
1892 and 1894. In the case of V the value of R'/R for 1893, though less remarkable, 
was below-the mean. In both H and V the values of R' for 1892 and 1894 were 
much in excess of that for 1893. For 1892 this excess was 21 per cent, for H, and 
33 per cent, for V. 
Table XLI. shows the order in which the years stand when placed in descending- 
order of range. Data for D are included to make the survey more complete. The 
