332 DR. C. CHREE: ATMOSPHERIC ELECTRIC POTENTIAL RESULTS AT KEW 
Table XVII.—Barometric Pressure. Diurnal Inequality (Unit O'OOOl inch). 
Range. 
Sum of 24 
differences. 
Day fall. 
Night fall. 
January . 
February . 
March. 
April. 
May. 
June. 
July. 
August. 
September. 
October. 
November. 
December. 
280 
262 
298 
320 
339 
307 
282 
328 
345 
250 
284 
295 
1655 
1657 
1840 
2101 
2157 
1933 
1747 
1852 
2052 
1822 
1561 
1649 
280 
262 
298 
320 
328 
307 
268 
328 
345 
250 
284 
295 
187 
176 
166 
166 
138 
128 
157 
93 
151 
210 
147 
188 
Arithmetic mean . . . 
299 
1835 
297 
159 
From mean diurnal"! 
268 
1699 
268 
152 
inequality for year J 
values in the summer months. The annual variations in the sum of the differences 
for B and P are shown together in fig. 1. To get a smooth curve in either case would 
probably require some 30 years’ observations, but the differences between the two 
curves speak for themselves. 
§ 29. Whilst the difference of phase in the two curves of fig. 4 is what most readily 
catches the eye, another difference seems not less important. In the P curve the 
afternoon maximum and the morning minimum are respectively the largest and 
smallest values in the day; but in the B curve it is the morning maximum and 
afternoon minimum that are the extremes. 
Comparing the two last columns in Table XVII. with the day and night falls of P 
in Table IV. we note again an essential difference. In the case of B the day fall is 
the larger in every month of the year. The excess, it is true, is most conspicuous in 
summer (the season when the day fall in P exceeds the night fall), but taking an 
arithmetic mean from the four winter months, November to February, we have 
day fall / night fall = 1'6 in B, 
= 0-55 in P. 
§ 30. To disclose more exactly the nature of the differences, the monthly and 
seasonal diurnal inequalities of barometric pressure were analysed in Fourier series 
with 24, 12, 8, and 6-hour terms. The results appear in Table XVIII., the notation 
used corresponding exactly to that of Table V. 
