1840.] On Lightning Conductors to Powder Magazines. 297 



nearest part of the conductor, which was twenty-six feet above 

 the roof— the distance being less than the simple height. The 

 parts struck contained a metallic cramp. 



13. Dr. Winthorp, of New Cambridge, reports, that a tree 

 was struck by lightning, when but fifty feet from a conductor 

 attached to the steeple of a church, which may reasonably be 

 supposed to have been at least fifty feet higher than the 

 tree.* 



14. All that we are entitled to infer from the facts before 

 us, is — that in order to give safety from direct and ordinary 

 discharges, we must erect so many conductors, that no point 

 of the roof shall be further from the conductor than twice the 

 length of the height of the conductor above the level of the 

 roofs ; and this applies only to flashes from clouds in a calm 

 atmosphere, and above the building. The area of protection is 

 unquestionably much contracted, under the circumstances, so 

 common in India, of a thunder cloud being blown with hurricane 

 velocity across a plain, before a furious squall. Nothing but 

 a line or chain of conductors connected together by horizontal 

 metal bars, and surrounding a building, can possibly protect it 

 from discharge under these paroxysmal storms. This is the 

 opinion I offered in my first report, dated the 27th December 

 1838, and I have now but to repeat, that one or even two 

 conductors are not an adequate protection ; and to ensure safety, 

 several must be erected. The subsequent considerations will 

 probably bear me out in repeating, that a properly built maga- 

 zine, with but one, or any inadequate number of conductors, is 

 in greater danger of explosion, than if it had none ; and that 

 with ever so many conductors, these should be placed at a con- 

 siderable distance from its walls. 



15. I proceed now in the attempt to sustain my opinion, that 

 "A magazine with but one, or any inadequate number of 

 conductors, is in greater danger than if it had none." 



* For details regarding the ease, see Annuaire for 1838. 



