326 DR. C. CHREE: ATMOSPHERIC ELECTRIC POTENTIAL RESULTS AT KEW 
rn 
1 ABLE 
XV. 
Excess from groups of clays of highest 
barometric pressure. 
Excess from groups of days of highest 
temperature. 
Baro¬ 
metric 
pressure 
in 
inches. 
Mean 
poten¬ 
tial. 
Poten¬ 
tial 
range. 
Non-cyclic effect. 
Tempe¬ 
rature, 
Fahren¬ 
heit. 
Mean 
poten¬ 
tial. 
Poten¬ 
tial 
range. 
Non-cyclic effect 
Alge¬ 
braic. 
Nume¬ 
rical. 
Alge¬ 
braic. 
Nume¬ 
rical. 
December . 
+ 0-447 
+ 36 
+ 52 
+ 2 
+ 51 
o 
+ 8-17 
- 108 
-46 
+ 6 
-29 
J anuary 
+ 0-540 
+ 59 
+ 10 
+ 4 
+ 4 
+ 7-76 
80 
+ 1 
+ 39 
-21 
June. 
+ 0-204 
+ 6 
- 7 
+ 13 
- 7 
+ 7-04 
- 18 
-20 
_ o 
- 18 
July. . . 
+ 0-205 
- 2 
- 7 
+ 16 
0 
+ 7-15 
+ 6 
+ 11 
+ 2 
- 11 
August . 
+ 0-252 
+ 13 
+ 11 
+ 31 
+ 9 
+ 5-44 
- 10 
- 9 
- 12 
+ 5 
§ 21. Instead of the mean value of P, or of the range of P, one may take for the 
criterion in grouping the selected days the value of some one meteorological element. 
This is a delicate mode of analysis, so far as the particular element is concerned ; but 
to carry it out for each element, for each month of the year, would have involved a 
very serious amount of arithmetic. The method has thus been applied only to some 
of the elements, and only for a selection of months. 
Table XV. gives the results thus found in the case of mean barometric pressure 
and mean daily temperature. December and January were selected as the months 
when according to Tables XIII. and XIV. the most decisive results might be expected. 
June, July and August were selected partly as representing summer, and partly in 
consequence of the exceptional nature of the July results in Tables XIII. and XIV. 
There appears in Table XV., as in the tables just mentioned, a decided association 
of high potential and large range with high barometer in December and January. 
In summer the range of barometric pressure is much less, so we should not in any 
case expect the effects on the potential to be so conspicuous. The results, however, 
are not merely smaller numerically in the three summer months, but they differ in sign 
amongst themselves. Table XV. thus agrees with Tables XIII. and XIV. in pointing 
to the conclusion that any association of mean potential or potential range with 
barometric pressure in summer is small and doubtful. In all five months there 
is in Table XV. an apparent tendency for P to increase more (or diminish less) 
throughout the 24 hours when pressure is high than when it is low ; curiously, how¬ 
ever, this effect seems much larger in the summer than in the winter months. The 
exceptionally large size of the non-cyclic effect, when taken irrespective of sign, in 
December when pressure is high, and the correspondingly large size of the range, are 
both probably due in considerable measure to fog. This is an element whose intensity 
at two consecutive midnights is usually widely different. 
