306 



ATMOSPHEEIC ELECTEICITY. 



The electroscope reading was noted when three successive points were registered by the electrometer 

 The electrometer registered every two minutes and it was easy to see whether there was an 

 appreciable leak from the change in the electroscope reading during the interval. The whole 

 system was then connected to earth for ten minutes or so to mark the zero line. Thus 

 on the trace itself were frequently recorded, the state of the insulation, the position of the 

 zero, and the deflection con-esponding to a known potential. 



It was necessary to find the reduction factor for converting the voltages recorded by the 

 electrometer to the potential gradient in volts per metre over a level surface. The frozen sea 

 within a few yards of the collector made this an easy matter. The necessary observations 

 were made on five days during November, December and January, ] 911-12. The reduction 

 factor proved to be linear over the whole scale which was a great convenience in working 

 up the records.* 



Potential Gradient and Drift.— When the air was full of drift snow the potential giadient 

 was as a rule very high. Also a very little drift would increase the potential above the 

 extreme range of the electrometer, i.e., 367 volts per metre. This is most clearly seen 

 when the average potential during different wind strengths is examined. When the observa- 

 tions for the whole year are arranged according to wind strength we get the following 

 result : — 



Table 156. 

 Potential Gradient and Wind. 



It will be seen that with winds of to 10 miles an hour the potential gradient was 

 104 volts per metre, while it steadily increased to > 304 volts per metre with winds over 

 forty miles an hour. The range of the electrometer would only record 367 volts per metre, 

 hence the value of the higher potentials coidd not be recorded. ^Vhen the needle reached 

 the extreme for only a shoi-t time in an hour, one could form some idea from the shape 

 of the existing curve what would have been the shape of the missing portion, and in this 

 way a rough estimate of the mean potential in the hour could be determined. When the 

 needle was against the extreme stop for an appreciable time, it was only possible to say that 

 the mean potential during the hour was > 367 volts per metre. In taldng the mean value 

 of the potential the sign > had to be neglected, hence all the values of the potential 

 gradient for wind strengths over 11 miles an hour are smaller than they would have been 

 if the true high potentials had been measured. Thus the sign > has been prefixed to the 

 values in table 156. It is clear that the highest mean potential which could be recorded 

 in this way is 367 volts per metre, and it will be noticed that this value is nearly reached 



* A description of the method of obtaining the reduction factor and of many other details of tlie apparatus 

 used in the atmospheric electricity work will be found in my paper entitled ' Instrumente zur Beobachtung der 

 atmospharischen Elektrizitat ' published in Physikalische Zeitschrift, Volume 14, page 41, 1913. 



