330 
Proceedings of the Royal Society of Edinburgh. [Sess. 
very small. Along with this curve is plotted the time of sunrise for the 
fifteenth day of each month. It will be seen that if a smooth curve be 
drawn through the points representing the values of cq for the various 
months, this curve is approximately parallel to that for sunrise, the 
difference being on an average about 20°-30°, or about two hours, so that 
we may say that roughly the first harmonic lags about two hours behind 
the sun. 
The phase of the second harmonic does not seem to undergo any 
very marked annual variation. Exceptionally low values are obtained in 
January and February, while the value for November is high compared 
with the others. For all the other months the value of a 2 varies com- 
paratively little, while the value of a 2 for the year works out at 183° — a 
value which agrees closely with that obtained at other stations, as will 
be shown later. 
§ 6. The harmonic analysis of the curves gives a more convenient 
method of comparing results at different stations than merely examining 
the numbers and the curves themselves. 
Table IV (a). 
Station. 
Period. 
Year. 
Winter. 
Summer. 
a 0 . 
aja 0 . 
aja 0 . 
aja 2 . 
CLi/(Iq. 
®2Ah>* 
^t/^2* 
aja 0 . 
a 2 /a 0 . 
aja 2 
Kew 
1898-1904 
159 
•053 
•156 
0-34 
•102 
•103 
0-99 
•079 
•213 
0-37 
Paris (B.C.) . 
1894-1898 
•220 
•104 
2-12 
•130 
•200 
0-65 
Paris (E.T.) . 
1896-1898 
•133 
•085 
1-56 
Karas j ok 
1903-1904 
139 
•281 
•165 
1-70 
•356 
•144 
2-46 
•165 
•093 
1-78 
Potsdam 
1904 
•269 
•101 
2-66 
•096 
•152 
0-63 
Kremsmiinster 
1902-1910 
106 
•243 
•136 
1-79 
•298 
•128 
2-33 
•195 
•152 
1-28 
Edinburgh 
1912 
167 
•142 
•102 
1-39 
•214 
•065 
3-30 
•086 
•150 
0-57 
Tables IV (a) and IV (b) give the values of the different amplitudes and 
phases obtained at various stations. We adopt the scheme given by Chree,* 
which is to express the amplitudes as fractions of the mean value of the 
potential gradient for the corresponding season of the year. In some cases 
the data are for all days, in others for quiet days, and the terms “ winter ” 
and “ summer” refer to periods of four or six months. The third column of 
this table gives the mean values of the potential gradient in volts per metre 
obtained at the various stations. The mean value for 1912 at Edinburgh 
is 167 volts per metre. 
* Art. “Atmos. Elect.,” Ency. Brit . , 11th ed., 1910. 
