312 DR. C. CHREE: ATMOSPHERIC ELECTRIC POTENTIAL RESULTS AT KEW 
The annual term is, in all cases, the more important of the two. The relative 
importance of the semi-annual term is much greater for the day fall than for the 
other elements; it is least for the mean value for the day. 
Table VI.—Annual Variation. Amplitudes and Phase Angles. 
1 
Pi- 
0i. 
P 2 . 
d,. 
Pi/M. 
Pa/M. 
P 2 /P 1 . 
Mean value for the day .... 
63-8 
o 
81 
7-49 
o 
84 
0-40 
0-05 
0-12 
Sum 24 hourly differences . 
92-0 
44 
26-7 
315 
0-20 
0-06 
0-29 
Range. 
15-7 
37 
2-80 
358 
0-23 
0-04 
0-18 
Day fall. 
11-3 
342 
8-95 
267 
0-22 
0-18 
0-79 
Night fall. 
15-9 
70 
6-49 
15 
0-26 
0-11 
0-41 
r x (diurnal inequality) .... 
8-60 
59 
2-61 
91 
0-60 
0-18 
0-30 
( i> » ) ... . 
8-44 
53 
3-26 
294 
0-33 
0-13 
0-39 
Comparison ivith Previous Results for Kew. 
§ 10. In the ‘ Phil. Trans.’ for 1868, pp. 347-361, atmospheric electricity data from 
Kew for a 2-year period (June, 1862, to May, 1864) were discussed by Professor 
J. D. Everett. The data were from all days of the year. Everett calculated 
Fourier coefficients corresponding to our c 1} c 2 , a u a 2 for the mean diurnal inequality 
of each individual month, and for each of the two years, and then for the 12 months 
and the year from the two year’s data combined. His amplitudes were expressed in 
terms ot an arbitrary unit. The results for corresponding months of the two years 
vary somewhat largely, and there is considerable difference between the mean results 
tor the two years. Thus even the values from the two year’s data combined, at 
least those for individual months of the year, are clearly exposed to rather large 
uncertainties. 
It should be noticed that Everett’s hour angles (loc. cit., p. 358) are counted from 
noon. Thus, in comparing his results with mine, 180° must be added to his E x 
(my «j); his E 2 and my a 2 are immediately comparable. The following remarks 
apply to his values for the two years combined :— 
My value for a x is above Everett’s from September to February, the mean excess 
being 18° P, and below his from March to August, the mean deficiency being 27° 10'. 
In the mean diurnal inequality for the year, however, Everett’s value for a x exceeds 
mine by only 1 6', corresponding to 4'4 minutes of time. As regards a 2 , my value is 
above Everett’s in six months, viz., May and July to November, the mean excess 
being 11° 7', and below his in the other six months, the mean deficiency being 12° 40'. 
In the mean diurnal inequality for the year his value of a 2 exceeds mine by 2° 10', 
answering to 4‘3 minutes of time. The differences in the phase angles in the two 
investigations, though comparatively small, especially in the case of the 12-hour 
