Mr. R. S. Brougli 07i a Case of Lightning. 107 



tive discharge may take place from the protector itself, which 

 under these circumstances becomes merely a source of danger*. 

 This tendency to disruptive discharge is taken advantage of 

 to protect telegraph-instruments from lightning. An earth- 

 wire is brought very near to the line-wire, from which it is 

 insulated by only a very thin stratum of air ; when the poten- 

 tial of the line-wire rises abnormally, a disruptive discharge 

 takes place at this point, and the receiving-instrument is thus 

 saved. 



I have twice lately seen it stated that Sir W. Thomson 

 found that the resistance of air to disruptive discharge de- 

 creased as the thickness of the stratum increased ; and a 

 French writer has referred the possibility of the occurrence 

 of lightning discharges several kilometres in length to this 

 cause. Sir W. Thomson's earlier experiments certainly showed 

 this unexpected result, probably due to the minute distances 

 at which he was operating ; but a later series of experiments, 

 made at larger distances, showed this result in a much less 

 marked degree ; and Sir W. Thomson himself says, ^' it seems 

 most probable that at still greater distances the electromotive 

 force will be found to be sensibly constant, as it was certainly 

 expected to be at all distances" j. 



Another assertion of the text-books is that the metallic rods 

 now employed as lightning-protectors on buildings do not 

 '' attract " lightning. This statement is literally true, accord- 

 ing to the meaning of the word " attract," but is untrue in 

 effect ; for such a rod lightning-protector determines a line 

 of maximum induction, and a discharge is more likely to occur 

 at the place than if the protector were not there. Professor 

 Clerk-Maxwell does not appear to hold this opinion ; but it 

 seems to me unquestionable that if a charged thunder-cloud, 

 driving before the wind, is carried over a building furnished 



* It is very necessary, therefore, that all systems of lightning-protec- 

 tors should be tested for resistance from time to time. Mr. Schwendler's 

 method of quantitatively testing "earths" has already been described 

 before the Society (Journ. As. Soc. of Bengal, part 2, vol. xl. 1871). 

 In this method two temporary auxiliary earths are required. Calling 

 the resistance of the lightning-discharger earth .r, and that of the auxiliary 

 earths respectively y and z, the three resistances x-{-y=a, :v-{-z=b, and 

 y-\-z—c are measured by any accurate method most convenient (e. g. 

 Wheatstone's bridge, differential galvanometer, tangent- or sine-galvano- 

 meter, &c., or even an empirically calibrated galvanoscope), the mean of 

 positive and negative readings being taken to eliminate any natural elec- 

 tromotive force between the earths. From the results thus obtained the 

 unknown resistance x can be calculated by the formula 



a-{-h — c 

 t Papers on Electrostatics and Magnetism, p. 259. 



