FROM SELECTED DAYS DURING THE SEVEN YEARS 1898 TO 1904. 
301 
If P : be the potential shown by the portable electrometer, and P 2 that shown by 
the electrograph at the same instant, P : -f- (D465 P 2 ) is a value for the factor by 
which the potential shown by the electrograph is to be multiplied to convert curve 
potentials into volts per metre of height in the open. Observations are taken with 
the portable electrometer nearly every day when the weather allows, and a mean 
value for the factor is calculated from each month’s comparisons. 
§ 3. When potential changes are very sudden, readings from the electrograph are 
doubtless partly dependent on its electrical capacity, and when the changes are very 
large, part of the trace is pretty sure to be lost. These large and sudden changes are 
mostly confined to times of rainfall, being specially prominent during thunderstorms. 
Further, during rain, drops are falling from all parts of the efflux tube, so that its 
record may not be solely, or possibly even mainly, determined by the jet at the end. 
For these and other reasons, it was decided to confine the tabulation of the 
electrograms to days when there was no rainfall, and to be content with 10 days a 
month. In selecting the 10 days one aimed at distributing them throughout the 
month, avoiding days in which negative potential was recorded, or there was reason 
to suspect the insulation. In some months there was a considerable excess of days to 
choose from, but in others it was difficult, and in a few cases impossible, to get as 
many as 10 suitable days. In one very wet month, in order to secure even 8 days, it 
was necessary, in two or three instances, to take in place of the natural day successive 
periods of 24 hours, comprising parts of two successive days. With this exception, a 
day always meant 24 hours, extending from midnight to midnight, G.M.T. All the 
data, except in Table IX., refer to Greenwich time, but the difference from this of 
local time is only lj minutes. 
Since 1902 the Annual Reports of the Observatory Department of the National 
Physical Laboratory have contained tables showing the mean monthly diurnal 
variation of the potential gradient from the selected days. Potential gradient is an 
exceptionally fluctuating element, and measurements of the actual photographic trace, 
taken at exact hours, may be far from representative of the mean value during 
60 consecutive minutes. The practice adopted has been to draw in pencil a continuous 
curve, following as nearly as the eye can judge the general trend of the trace, and to 
measure the ordinates of this curve at exact hours, G.M.T. 
During the past two years, the data for earlier years have been similarly treated, 
and the present paper deals with the results from the seven years 1898 to 1904. 
Mean Monthly and Annual Values of the Potential Gradient. 
§ 4. Table I. gives the mean value of P (the potential gradient in the open in volts 
per metre) for each individual month and for each year, and likewise the mean annual 
inequality based on the seven years’ results. The annual inequality is exhibited 
