60 
DR. S. CHAPMAN ON THE SOLAR AND LUNAR 
From Table C the amplitudes C\_j may be calculated for the solar diurnal 
magnetic variation. The relative amplitudes are nearly the same in 1905 and 1902, 
as the following table indicates :— 
Table If.—Ratios of C" J+1 in 1905 and 1902. 
C 2 l 
Cs 2 - 
c 4 8 . 
c s . 
Equinox. 
1-43 
1-31 
1-25 
1-35 
Solstice. 
1-47 
1-31 
1 • 32 
1 60 
The increased amplitudes all round, in years of sunspot maximum, naturally point 
to a general increase in the electrical conductivity of the atmosphere, with little 
or no change in its functional dependence on the sun’s zenith distance. 
The lunar magnetic data refer to the quieter years of a solar cycle, and may be 
compared with the solar data for a mean of 1905 and 1902, giving double weight 
to the latter year. The comparison leads to the following results (the tenfold unit 
in Table F, as contrasted with Table J, should be remembered):—- 
Table S.—Ratios of Solar and Lunar E". 
CA 
Cat 
c 4 3 . 
C 5 4 . 
Equinox. 
21-8 
10-4 
10-0 
7-6 
Solstice. 
18-5 
10-2 
10-0 
7-3 
The ratios for the last three harmonics are in moderate agreement, especially in 
view of the considerable seasonal variations in the amplitudes of the third and fourth 
harmonics. If the ratios for the first harmonic had also been about 10, little doubt 
could remain that all four are produced by a single semi-diurnal atmospheric 
oscillation, roughly of type Q 2 2 , as in the case of the lunar diurnal magnetic 
variations. 
Even as it is, the fact that the harmonic Q 2 _1 will appear with Q 2 X in the solar, but 
not in the lunar data, leaves a possibility of explaining the above figures without 
introducing a further atmospheric oscillation into the theory. The agreement of 
phase, and the apparently more fundamental character of the 12-hour oscillation as 
compared with that of the 24-hour period (at the earth’s surface) favour this solution. 
It is worth while, however, to examine the magnetic effects of a 24-hour atmo¬ 
spheric oscillation. Considering the amplitudes of the magnetic components, on the 
