286 On Lightning Conductors to Powder Magazines. [No. 99. 



(No. 5.) 



Mr. Faraday's letter to Mr. Secretary Melvill, on 

 the preceding papers. 



Roy at Institution, 5th September, 1839. 



Sir, — I have the honor to acknowledge your letter and the 

 papers, and having read the latter, beg leave to state that my 

 opinion is in favor of lightning conductors. It is no doubt true 

 that low rounded buildings, such as I understand the powder 

 magazines in the East Indies to be, are but very little liable to 

 be struck by lightning, but then if they are struck, the destruc- 

 tion and injury may be very great. It is also, I think, very 

 probable that a lightning conductor may, under certain circum- 

 stances, cause an electric discharge to take place where none 

 would have occurred no conductor being present, though, on the 

 other hand, there is some evidence to show that conductors cause 

 a diminution in the number of electric discharges to the earth 

 at a given place. It is also very certain that a badly erected 

 conductor is worse than none, and may cause great injury. But 

 notwithstanding these considerations, I have the strongest con- 

 viction in my mind that conductors welt applied are perfect 

 defenders of buildings from harm by lightning. Dr. O'Shaugh- 

 nessy's papers are very valuable, and serve to confirm my pre- 

 vious impressions ; but it would be impossible for me to go over 

 the whole of the opinions and evidence sent me, without at the 

 same time going into a far greater mass dispersed here and there. 

 I would rather refer you at once to M. Arago's popular view of 

 the subject in the Annuaire for 1838, pp. 221, 549, &c. with 

 which I, in almost every point, agree. 



I would certainly recommend copper conductors instead of 

 iron, for the former metal conducts electricity almost seven times 

 better than the latter. When struck, it not only conducts the 

 shock much better, but in the pre-determination of the stroke it 

 determines more of the electricity to itself than otherwise would 

 fall upon it, and therefore tends in any case of a divided shock 

 to leave less to fall elsewhere in its neighbourhood. 



I should prefer them pointed. I should not put them far from 

 the building at their upper extremity, or in their course down- 



