374 Mr. 0. Toralinson on Lightning 



after such an explosion would throw much, if any, light upon 

 the cause of the accident. As to the inspection of the conduc- 

 tors shortly before the explosion, it is necessary to see the report, 

 if any were made, and also to know whether the inspectors 

 were competent to undertake the task. 



In the case of the Purfleet magazines, Franklin's arrangement 

 of the protecting conductors was judicious, hut the magazines 

 themselves were faulty in construction. The five magazines 

 were placed side by side with interspaces of about 57 feet. If 

 an accident had happened to any one, the others would probably 

 have shared its fate. The barrels of gunpowder were piled on 

 each other up to the spring of the arches, and each barrel was 

 bound with four copper hoops, while a number of iron bars 

 passed through the arches to support the timbers on which a 

 crane was worked, thus forming broken conductors within 

 the building. These iron bars were, however, removed at the 

 suggestion of some Fellows of the Royal Society, with the 

 approval of the Committee. 



Snow Harris admitted the difficulty of providing complete 

 protection to powder-magazines, from the fact that they are 

 frequently constructed in a series of long straggling buildings, 

 and hence are more likely to be struck at some point distant 

 from the conductors. He recommended as the safest form 

 for a powder-magazine a circular building with a conical 

 metallic roof, furnished with a projecting pointed rod from 

 the vertex, the cylindrical body being furnished with several 

 vertical metal rods, attached to the roof and passing into 

 damp ground or water. If such buildings were covered on 

 the outside entirely with metal, still greater security would be 

 attained. 



The late Professor Clerk-Maxwell was so kind as to com- 

 municate to me his ideas on the same subject. Under his 

 system the projecting point and the ground connexion were 

 dispensed with, the function of the point being rather to tap 

 the thunder-cloud, or the atmospheric charge, than protect the 

 building ; whereas the object to be attained should rather be 

 to prevent the possibility of a discharge taking place within a 

 certain space, such as a gunpowder manufactory or magazine. 



His system is based upon the postulate that an electric dis- 

 charge cannot take place between two bodies unless the differ- 

 ence of their potentials be sufficiently great, compared with 

 the distance between them. If we can keep the potentials of 

 all bodies within a certain space equal, or nearly so, no dis- 

 charge will take place between them. We may secure this by 

 connecting all these bodies by means of good metallic con- 



