Current and on the Origin of Atmospheric Electricity. 367 



We have obtained by these observations : 



1. The charge on our test-plate when earth connected and 



exposed to the earth's electric field. This is directly 

 proportional to the potential gradient, and if we 

 always fix the instrument at the same height above 

 the ground we may determine once for all the factor 

 by which to obtain the potential gradient from the 

 charge on the test-plate. This method of measuring 

 the potential gradient has the advantage of being 

 practically instantaneous and of not requiring the use 

 of any flame, radio-active substance, water dropper or 

 other collector. 



2. The charge which has passed from the atmosphere into 



the test-plate in a definite number of minutes when 

 the test-plate is kept at the same potential as if 

 earth connected. 



3. The ratio of the charge passing per minute into the 



test-plate from the atmosphere (obtained from 2) to 

 the charge on the test-plate (1) both measured by 

 the compensator may be taken as giving the dissipa- 

 tion constant under conditions of exposure identical 

 with those to which natural earth -connected bodies 

 are exposed. 

 It is only for a few seconds at the beginning and end of the 

 measurement, in the interval between taking off or putting on the 

 cover and bringing back the gold-leaf to its zero position by 

 the compensator, that the potential of the test-plate and the 

 connected system differs by more than a very small fraction of 

 a volt from the potential when earth connected. The danger of 

 error due to faulty insulation of the test-plate is therefore very 

 small. The insulation can however be readily tested by putting 

 on the cover and charging the test-plate system to any desired 

 potential by means of the compensator; the insulation was always 

 found to be high when thus tested. A failure of the insulation of 

 the quartz Leyden jar supporting the inner case of the electro- 

 meter would be more serious; it was found however to remain 

 practically perfect, the change of zero only amounting to a small 

 fraction of a volt in 24 hours, and the system not requiring to 

 have its charge replenished for some months when once charged. 

 There remains to be considered the working of the compensator. 

 This consists of a brass tube, kept at a constant negative potential 

 by being connected to a quartz Leyden jar, and forming with an 

 inner concentric rod attached to the terminal of the gold-leaf an 

 air condenser of variable capacity. The adjustment always being 

 made in such a way as to bring the gold-leaf to zero potential, 

 the charge held on the rod was proportional to the potential of the 



