Current and on the Origin of Atmospheric Electricity. 375 



test the zero again. Finally the compensator was drawn out, the 

 earth connection broken, and the first operation of finding the 

 charge on the exposed test-plate at zero potential repeated, so 

 that any change of the potential gradient might be detected. 

 Generally the change in the potential gradient between the two 

 measurements of charge was small, and the mean was taken as 

 the charge for the time during which the current was measured, 

 and this value was used in deducing the dissipation factor. A 

 second measurement of charge and current could now be made as 

 soon as might be desired. The testing of the compensator constant 

 did not require as a rule to be made again except at the end of 

 the whole series of observations ; it generally showed no appreciable 

 change in a few hours. 



The results of the hill-top measurements are given in the table 

 on page 376. 



The time is that at which the cover was removed at the 

 beginning of an observation. It was replaced from two to five 

 minutes later. 



The charge and the leak are those of the test-plate, and are 

 given in electrostatic units ; I have not attempted to deduce the 

 corresponding charge and current per square cm. of the ground 

 The last column, giving the ratio of the leak per minute to the 

 charge, may be taken as the dissipation factor measured under 

 conditions very like those holding at the surface of the ground. 

 The observations were taken under very varied conditions. The 

 smallest dissipation, of the order of one per cent, per minute, was 

 observed on April 10th on a cloudless day with a somewhat thick 

 haze. Values equal to or approaching 10 per cent, per minute 

 were not unfrequent. Fairly high values exceeding 5 per cent. 

 were obtained on the only occasion when such measurements were 

 attempted at night (April 8th, between 11 p.m. and midnight). 

 The night was perfectl}* clear and calm ; the observations were 

 made without the use of any artificial light which might have 

 influenced the leakage; they were carried out entirely by the 

 light of the full moon. 



The mean value of the dissipation factor from these observations 

 is 5*6 per cent, per minute. 



Objections might with some justification be raised that the 

 dissipation constant as deduced by the present or other methods 

 can hardly be applied with confidence to the calculation of the 

 actual earth-air current, the materials covering the earth's surface 

 being very different from the metal test-plate. 



The matter is very readily tested with the apparatus which 

 has been described. For any conducting body may be placed on 

 the test-plate, and the charge upon it and current through it, 

 when exposed to the earth's electric field and kept at zero 



25—2 



