FAEMEKS' INSTITUTES. 359 



or through the whole capacity of the rod has arisen principally from experi- 

 ments on galvanic electricity, which having little or no repulsive energy passes 

 through the whole substance of the rod, and also from experiments in which a 

 very large quantity of frictional electricity is transmitted through a small wire ; 

 in this case the metal is resolved into its elements and reduced to an impalpable 

 powder. 



In the case, however, of the transmission of atmospheric electricity through 

 a rod of sufficient size to transmit the discharge freely, there can be no doubt 

 that it tends to pass at the surface, the thickness of the stratum of electricity 

 varying with the diameter of the rod and the amount and intensity of the 

 charge. 



To test this by actual experiment I made the following arrangement : through 

 a gun-barrel about two feet in length a copper wire was passed, the ends pro- 

 jecting. The middle of the wire in the barrel was coiled into the form of a 

 magnetiziug spiral and the ends of the gun-barrel were closed with plugs of tin- 

 foil, so as to make a perfect metallic connection between the wire and the bar- 

 rel. On the outside of the barrel another magnetizing spiral was placed, the 

 whole arrangement being shown in the annexed sketch. 



A powerful charge was now sent through the copper wire from a Leyden jar of 

 about two gallons capacity. The needle within the barrel showed not the least 

 sign of magnetism, while the one on the outside was strongly magnetic. 



From this experiment, I conclude that a gas-pipe can convey an ordinary 

 charge of electricity from the clouds as well as a solid rod of the same diameter. 



The repulsive energy of the electrical discharge at right angles to the axis 

 remains of the same intensity as in the case of a statical charge. This I have 

 shown to be tlie case by drawing sparks of considerable intensity from a con- 

 ductor, one end of which was connected with the ground while sparks Avere 

 thrown on the other end from a large prime conductor. This spark is of a 

 peculiar character, for though it gives a pungent shock and sets fire to combus- 

 tible substances, such as the electrical pistol, it does not affect a sensitive gold 

 leaf electrometer. The fact is, it consists of two sparks, the one negative and 

 the other positive. The rod during the transmission of the electricity through 

 it is charged + at the upper end and immediately in advance of this point it is 

 charged — by induction, and in the discharge the electricity passes through it 

 in the form of a series of + and — waves. 



Yours very truly, 



Pkof. R. C. Kedzie, JOSEPH HENRY, 



Lansing, Michigan. Sec. Smithsonian Institution. 



Smithsonian Institution, \ 

 Washington, D. C. Ajml 15, 1876. \ 



R. C. Kedzie, Lansing, Midi. : 



Deae Sir: — Your letter wns received by due course of mail, but a press of 

 business connected with the preparation of the annual report for 1875 and the 

 lighthouse board has prevented an earlier reply. 



I now have to say that as far as I know I am the only person who has made 



