FROM SELECTED DAYS DURING THE SEVEN YEARS i898 TO 1904. 
term, are too systematic to be ascribed to chance, but several explanations are 
possible. 
Everett’s table on his p. 358 agrees with my Table V. in making c x (his A,) 
greatest in December and January, and least m July and August; and his difference 
between the maximum and minimum values is even greater than mine. Also his 
largest values for c 2 (his A 2 ) occur, like mine, in February and March, his least values 
occurring in May, June, and November. There is thus a fairly close general 
resemblance between our results for the amplitudes. There is, however, a very 
important difference as regards the relative importance of c l and c 2 . We have, in 
short, 
Everett (1862-4). Chree (1898-1904). 
arithmetic mean of 12 monthly values of c 9 
_ _~ _ 0 * S 5 1*75 
arithmetic mean of 12 monthly values of c\ " 
c, in mean diurnal inequality for th e year 
c x in mean diurnal inequality for the year 
2-95 
Some further aspects of this difference are considered later in § 13. 
Everett also expressed the annual variation in the mean daily potential gradient 
as a Fourier series with annual and semi-annual terms. His times were measured 
from the middle of January, so that his phase angles (E x and E 3 in his notation) have 
to be diminished accordingly for comparison with mine. His results from the two 
years combined are in the notation of Table VI., counting time from the beginning of 
O o O 
the year, 
K 0. 2 . Pa/M. Po/M. Po/Pa. 
90° 249° 0-36 (P02 0'05 
Everett’s results for the annual term are thus fairly similar to mine, but he makes 
the amplitude of the semi-annual term even less than I do, and his value for its phase 
angle is quite different from mine. This is not, however, very surprising, as the 
phase angles for his two years taken separately differ by 100°, and the amplitude for 
the second year is only f of' that for the first. Also inspection of Everett’s annual 
variation curve in his plate XXI. shows that its representation by only annual and 
semi-annual terms must be very rough. 
Comparison with Chauveau’s Results for other Stations. 
§ 11. The intercomparison of results from different stations is difficult owing to 
uncertainty as to the real significance of the units employed. The length and the 
height above ground of the discharge tubes of water droppers vary greatly. Some 
are in low, others in lofty buildings, so that even if all were of one length and at 
one height they would not give comparable results. 
To meet this difficulty so far as possible, in the article on “Atmospheric Electricity” 
VOL. CCVI.-—A. 2 S 
