Daily Range of Atmospheric Potential Gradient. 11 
by means of a battery of small Leclanche cells. The E.M.F. of the battery 
was tested by a voltmeter from time to time, the number of cells in use 
being varied so as to keep the potential constant. 
The instrument was placed in a small house built at some distance 
(380 yards) from other buildings and trees. A hole in the wall permitted 
a well-insulated rod, to which the radium collector was attached, to project 
about 1 metre away from the wall. The collector was a plate of brass, 
coated with a thin layer of an impure radium barium bromide, containing 
not more than 1 per cent, of radium. The radium salt was protected by 
a layer of shellac varnish. This disc was covered by a conical cap of 
thin copper to prevent rain falling directly upon it ; during the time the 
experiments lasted it preserved unimpaired its efficiency, as shown by 
testing the intensity of the /3 and y radiation. The presence of the 
building considerably disturbed the normal potential gradient, but a 
correcting factor was readily found by taking other observations at some 
distance from the building, using a fine wire stretched from insulators 
attached to two vertical poles. This factor was 3, and this number 
has been used in reducing the results. The height of the collector 
above the ground was 2 metres. 
The observations were started on July 10th, and continued daily until 
December 30th. This period turned out to be a rather unfortunate one 
on account of the abnormal weather, and as a record of the normal 
variations in the potential gradient is not of much use, but as a means 
of throwing light on the cause of the variations of the potential gradient 
it is of some considerable importance. The results for each hour are 
given in the table, and in some cases of particular interest the record 
itself is reproduced. The normal course of the variations in the potential 
gradient seems to be of much the same type in all parts of the world — 
that is, there are two periods of maximum potential during the day ; these 
are followed by two long periods when the potential is low and fairly 
uniform ; but circumstances may arise which cause this course to be 
deviated from considerably. The record for August 8th is a fairly normal 
one. The maxima occurred at 7.30 a.m. and at 7 p.m., with the minimum 
at 1 to 2 a.m. ; the curve 1 shows the record. From July 10th to August 
1st the curves obtained are generally of the above character, but with 
rather smaller maxima than had been observed in previous years. The 
days were fairly clear, and it has been shown * that haze, which usually 
occurs during the winter months, is accompanied by considerable rise in 
the potential. In a Scotch mist, as long as the insulation held, the charge 
was generally too large to measure. 
It does not appear profitable to analyse the results or to discuss them 
with the idea of deducing any mean daily value of the potential, as is 
* South African Journal of Science, March, 1912. 
