NOTE ON A LIGHTNING DISCHARGE. 27 



beneath the surface of the ground, to reach charged pieces of 

 water, throwing out branches on their way to discharge the 

 intervening surface of the ground. 



When anything simihxr to this takes place on or near a 

 building on which there is a conductor, it is at once put down to 

 a bad earth-connection, while apparently it is due to the wet 

 surface of the ground being a better conductor than the stratum 

 lying between the subsoil and the charged surface. 



Cases in which large numbers of sheep, cattle or people 

 have been killed by lightning are easily comprehended, after 

 looking at this sketch. It is evident that had a number of 

 people or cattle been standing beneath these trees and occupying 

 the space covered by the furrows, that they would have consti- 

 tuted the charged surface and that the discharge would have 

 passed thi'ough their bodies instead of, as in this instance, 

 through the ground. 



The practical deductions to be drawn from this case seem 

 to be that all surfaces which are liable to become charged should 

 be put in communication with the system of conductors on a 

 building. In other words all roofs should be connected by their 

 ridges and gutters. While, as much as ])ossible, lightning con- 

 ductors should be arranged so as to be lightning distributors. 

 For this purpose all the conductors on a building should be 

 connected together, and the earth-plates so spaced round it that 

 the discharge may be spread out over as large a surface as 

 possible. 



