FROM SELECTED DAYS DURING THE SEVEN YEARS 1898 TO 1904. 
327 
In the case of temperature, July stands out from the other months exactly as it 
did in fables XIII. and XIV. The association of high temperature with low 
potential is again most conspicuous in the winter months. 
Accompanying the diminished value of P we have a diminution m the non-cyclic 
effect treated numerically. The apparent absence of temperature effect on the 
potential range in January, and the exceptionally large tendency to an increase in P 
during days when temperature was high, are probably due to some common cause. 
§ 22. If there were a direct causal connection between any one meteorological element 
and potential gradient or range, we ought to get the same or at least closely similar 
results for the ratio (difference of potential)/(difference of element) when the ratio is 
derived from a Table such as XV., and when it is derived from a Table such as XIII. 
or XIV. If such an influence though not unique were dominant, the two ratios so 
found should be at least of the same order. Combining,’ the results from the two 
months, December and January, I find for the ratios of the .differences—the units 
being for temperature 1°, for pressure 1 inch— 
Mean potential 
Potential range 
Mean potential 
Potential range 
Temperature 
Temperature 
Barometric pressure ’ 
Barometric pressure 
By Table XV. - 12 
— 3 
+ 96 
+ 63 
,, Tables XIII. and XIV.. . -23 
-38 
+ 1053 
+ 666 
The case of mean potential and temperature is the only one in which the ratios 
are of the same order. 
The mean temperature of the selected days in December and January was 38°'3 F. 
If instead of December and January we had taken June and August, we should 
have got altogether different results for the potential-temperature ratios, while taking 
July the sign even would have changed. Thus, if any direct relationship exists 
between potential and temperature, it must give a ratio between potential change 
and temperature change, which alters rapidly, and which changes sign at a tempera¬ 
ture between 64 0, 5 (the mean temperature for the selected days in July) and 60 o, 3 
(the mean temperature for the selected days of June and August). 
It may be the purest accident, but somewhat similar ideas present themselves when 
we consider the mean seasonal values of potential and temperature. For these, 
taking the temperatures of the selected days, I find—- 
Mean potential. 
Mean temperature. 
Winter. 
217 
39-2 
Equinox. 
148 
48-8 
Summer. 
111 
59-3 
