ATMOSPHERIC ELECTRICITY. 237 



tial gradient was now over 500 volts per metre, this persisting up 

 till 10.30 p.m. On the next day, the charge in the air vi^as prac- 

 tically nil until 10 a.m., when it became just measurable. 



Observ^ations were taken not far from the railway, and it 

 appeared that the passing of a train had a very considerable dis- 

 turbing effect, which was traced to the steam emitted from the 

 funnel or from the safety-valve, or even from that escaping from 

 the cylinder when blowing off. At 200 yards' distance the pass- 

 ing of an engine under steam would cause the gold leaf to be 

 deflected up to the maximum, that is, the potential gradient would 

 rise to over 550 volts per metre. 



The charge Avas always positive, and persisted in the air 

 for a long time after the cloud had suffered visible condensation, 

 in some instances for more than fifteen minutes. The charge 

 increased during the day, and on occasions went up to 350 volts. 

 The wind was rather strong, this accounting to some extent for 

 the variations. 



At Ladysimith, 6.30 to 9 a.m. : Potential varied from 

 40 to 100 volts. At various stations between Ladysmith and 

 Pietermaritzburg the potential varied only slightly, but as the 

 time of day varied no very satisfactory conclusions could be 

 drawn. At Durban the potential gradient was always greater 

 than at any point after Van Reenen. The weather was warm 

 and hazy, and the gradient never went beyond 160 volts per 

 metre, a value very nmch less than had been found at Cape Town. 

 On the return journey, a good set of observations was taken 

 between Ladysmith and Vian Reenen ; the potential gradient 

 falling as the altitude increased. 



These observations cannot claim any very high degree of 

 accuracy, but they are comparative, and appear to show that 

 the potential gradient does diminish with the altitude, and that 

 over a large stretch of coimtry. At appnoximately the same height 

 above sea-level the potential gradient has generally the same 

 value. The gradient could also be determined if observations 

 were taken of the conducting power of the air and of the 

 magnitude of the earth-air current. 



It would be equal to the quotient of the current of the con- 

 ductivity. Now the conductivity of the air increases with the 

 altitude owing to the diminution' in pressure, and perhaps also 

 owing to the more potent eft'ect of the ultra-violet light of the 

 sun causing more complete ionisation, from which it may be 

 inferred that the potential gradient will fall. 



The nature of the noTmal charge observed was always posi- 

 tive, Slave in the case of the Victoria Falls, where it was strongly 

 negative. It is well known that when water is broken up into 

 a fine spray by mechanical means, the water becomes positively 

 electrified, whilst the air in the neighbourhood becomes nega- 

 tively charged. The condensation of steam to water is evidently 

 accompanied by the production of a positive charge in the air; 

 hence it may be inferred that the drops of water falling to the 

 earth s'hould carry with them a positive charge. This requires 

 further investigation. 



