246 
DR. C. CHREE: DISCUSSION OF KEW MAGNETIC DATA 
Relatively considered, the amplitudes of the first four Fourier waves show a fairly 
similar range of annual variation. On the whole the range is largest in c x and least 
in c 3 ; but c 3 and c 4 , having a double annual period, vary most rapidly. 
Table XXV.—Ratios to Arithmetic Mean (ll years’ data). 
Month. 
c x . 
C2- 
c 3 . 
I 
Ci- 
H. 
v - 
Mean. 
II. 
V. 
- 
Mean. 
H. 
Y. 
Mean. 
H. 
y. 
Mean. 
January . . . 
0-29 
0-50 
0-40 
0-61 
0-29 
0’45 
0-85 
0-33 
0-59 
1-03 
0-72 
0-87 
February . . 
0-42 
0-71 
0-56 
0-64 
0-60 
0-62 
1-01 
0-77 
0-89 
1-01 
0-75 
0-88 
March .... 
0v84 
1-00 
0-92 
1-02 
1-18 
1-10 
1-41 
1-43 
1-42 
1-31 
1-58 
1-45 
April. 
1-33 
1-19 
1-26 
1 • 36 
1-43 
1-39 
1-36 
1-61 
1-49 
1-32 
1-32 
1-32 
May. 
1-46 
1'55 
1-50 
1-16 
1-68 
1-42 
0-47 
1-23 
0-85 
•0-83 
1-02 
0-93 
June. 
1-59 
1-47 
1-53 
1-27 
1-46 
1-37 
0-65 
0-83 
0-74 
0-57 
0-58 
0-57 
Jnly. 
1-59 
1-57 
1-58 
1-27 
1-41 
1-34 
0-76 
0-97 
0-87 
0-59 
0-60 
0-59 
August . . . 
1-48 
R05 
1-26 
1-14 
1-37 
1-25 
1-20 
1-45 
1-32 
1-15 
1-00 
1-08 
September . . 
1-22 
1-09 
1-15 
1-03 
1-07 
1-05 
1-45 
1-21 
1-33 
1-51 
1-31 
1-41 
October . . . 
1-01 
0-82 
0-91 
1-10 
0-77 
0-94 
1-38 
1-11 
1-25 
1-30 
1-68 
1-49 
November . . 
0-48 
0-64 
0-56 
0-82 
0-46 
0-64 
0-84 
0-58 
0-71 
0-91 
1-10 
1-01 
December . . 
•0-30 
0-42 
0-36 
0-60 
0-28 
0-44 
0-61 
0-46 
0-53 
0-46 
0-34 
0-40 
§ 21. Fourier coefficients were not calculated for the individual months of the year 
except for H and V. For the other elements they were calculated only for the 
diurnal inequalities from the seasons and the year. Table XXVI. compares the 
amplitudes in these seasonal diurnal variations for H, V, T, N, W, and I, use being 
made in dealing with N and W of the results previously obtained for D. In the case 
of c x and c 2 the equinoctial value is always intermediate in size between the winter and 
summer values, and somewhat in excess of the value for the whole year. In the case 
of c 3 and c 4 the equinoctial value is invariably in excess of both the winter and summer 
values, and much in excess of the value for the year. In the case of c 3 the summer 
value is the lowest for two elements, N and Id ; in the case of c 4 the winter value is 
less than the summer value in no element except I. 
In the case of all the Fourier waves for the three rectangular components V, N, and 
W, the year value of the amplitude is largest in W and least in V. This is in 
general true also of the three seasonal values, but in equinox c x is larger in N than in 
W, and there are one or two cases in which the V value is not the lowest. 
Table XXVII. shows the ratios borne by the amplitudes of the 12-, 8-, and 6-hour 
waves to the amplitude of the corresponding 24-hour wave. In the case of H, T, N, 
and I, the importance of the 12-, 8-, and 6-hour waves falls relative to that of the 
24-hour wave as we pass from winter to equinox, and from equinox to summer; but 
in W and V the relative importance of the shorter period waves is greatest in equinox, 
and in the case of the 12-hour wave it is least in winter. 
