180 DAVENPORT ACADEMY OF NATURAL SCIENCES. 



came the thunder crash. He was, of course, fully aware that the 

 mill had been "struck." Almost immediately there poured forth 

 from it a volume of smoke and then of flame, so that it seemed, as he 

 expressed it, as if "it must have made a l)i(j hole'" in tlie roof or 

 walls. 



We must probably be forced to the conclusion that the Chambers 

 rod is, like all lightning rods, not infallible; and the rod on the Phoe- 

 nix Mills must be placed in the category of failures. All experience 

 seems to show that no lightning rod aifords complete security against 

 violent lightning discharges. Their chief usefulness, probably, is as 

 equalizers, tending to the restoration of equilibrium of disturbed 

 electrical conditions, and thus to prevent, modify or weaken the dis- 

 ruptive discharges. In this view there seems to be no good reason 

 for assuming that the Chambers rod is useless. How man}^ light- 

 ning discharges have been prevented by lightning rods, or to what 

 extent they have been weakened, can of course, in the nature of the 

 case, never be known. 



It may be assumed that, to be at all efficacious even in this way, 

 the rod must have a ground connection, but this, as I l>elieve, in not 

 proven^ and some ex])eriments apparently point dii-ectly to the oppo- 

 site conclusion. 



AVe very well know that every sharp point or edge of a conductor 

 or a body in the condition of electrical tension, affoids an opportu- 

 nity for the escape of that tension, just as surely as that a hole in a 

 hose tilled with water at a high pressui-e allows a ))ortion to escape, 

 and thus diminishes the pressure. 



From all experiments and investigations, as well as from theoret- 

 ical considerations, it appeals that the best security possible would 

 be afforded to buildings by having extensire metallic surfaces, with 

 many sharp edges, points and corners, — the larger the surface, and 

 the more points and rough edges the better — whether connected 

 with the earth by conductors or not; and this latter coiulition )^i<n/ he 

 a matter of minor importance. 



There seems to me to be no reason to fear that the rod increases 

 the danger of injury by lightning under any circumstances. If it 

 has any effect it must be to diminish the danger in a greater or less 

 degree. Otherwise we must also condemn all metallic railiiig and 

 ornaments which are so common on the tops of buildings. 



