Theory of Thunderstorm Electricity. 



633 



considerable extent, which, if the storm is to last long, must 

 be little less than 2,500 metres in vertical thickness ; and, 

 finally, a second cloud-layer where water is condensed on 

 the negative ions, and the separation of electricity takes 

 place. 



From the time when the theory under consideration first 

 came to my notice I have watched thunderstorms with special 

 attention to the conditions which the theory presupposes. 

 In no single case have I seen what appeared to be a large 

 cumulus cloud with another cloud above it, but on the con- 

 trary, I have seen on several occasions distant clouds, from 

 which heavv rain associated with thunder and lightning was 

 falling, with sharply defined cumulus heads and quite clear 

 sky above. Both Wilson and Gerdien refer to the false 

 cirrus seen above thunder-clouds as very likely to be the 

 cloud connected with the condensation layer. But every 

 time that I have seen a layer of cirrus-cloud accompanying 

 a thunder-cloud it has appeared to me as if the top of the 

 cumulus-cloud had reached a part of the atmosphere which 

 had robbed it of its rounded form and spread it out into a 

 cirro-stratus cloud. I have never seen anything which would 

 lead me to believe that between the heavy cumulus-cloud 

 of the lower atmosphere and the cirro-stratus of the upper 

 atmosphere there was a large cloudless region. 



With regard to the electrical discharge of a thunderstorm; 

 as far as my own observations go I should certainly say that 

 the most violent and rapid discharges do not take place in 

 the upper atmosphere : on the contrary they have always 

 appeared to me to have their chief source in the centre of 

 the large cumulo-nimbus cloud from which the rain was 

 falling. The following account of a small thunderstorm may 

 be interesting from the point of view of this theory. During 

 an afternoon in August there had been heavy rain in Simla, 

 but the sky commenced to clear before six o'clock. Soon 

 after the sky had cleared so far as to leave the hills around 

 visible, with large masses of clouds still clinging to them in 

 places, a cloud in the form of an isolated pillar of compara- 

 tively small cross-section was seen to be rapidly extending 

 upwards. The cloud was evidently the result of a rapid rise 

 of moist air in a local current ; it extended upwards with a 

 beautifully rounded cumulus boundary at the top, and the 

 centre was seen to be illuminated at intervals of about a 

 minute by vivid internal lightning discharges. In an hour 

 the cloud had lost its shape and become an ill-defined mass 

 of cloud. In this case it could be said witli absolute certainty 

 that there were not two clouds separated by a supersaturated 



