FARMEES' INSTITUTES. 363 



Oa passing a curieufc of galvanic electricity from + to — , the needle of the 

 galvanometer was violently deflected, and a soft iron wire in the magnetizing 

 spiral was made magnetic. This experiment reproduces with (/a/ya^i^c electricity 

 the same apparent tendency to surface action, which was so strikingly exhibited 

 in Prof. Henry's experiment with frictional electricity. In this experiment no 

 one will attempt to explain tliis division of the current, and the passage of a 

 part of the electricity through the exterior conductor by any assumption of 

 repulsion which exists in the statical condition of frictional electricity, because 

 galvanic electricity is devoid of all repulsion. Will not the following proposi- 

 tion cover all the essential facts iu both experiments? viz. : the resistance to 

 the passage of a current of electricity being directly as the amount of elec- 

 tricity, and inversely as the conductive capacity of the conductor, and electricity 

 of the same kind having absolutely no cohesion, when two paths are open for 

 the passage of a current of electricity the current will divide and pursue both 

 paths — most by the path offering least resistance, and least by the path of 

 greatest resistance. 



Since the gun-barrel exerts such a controlling influence over the magnetizing 

 power of an inclosed spiral, the intensity of the magnetized needle in such a 

 spiral was discarded as a measure of the amount of electricity which traverses 

 such spiral, and some other more accurate means of measuring the relative 

 amount of electricity which passes by the interior and exterior circuits was 

 sought for. Failing in my efforts to construct electrical thermometers of suffi- 

 cient accuracy for this purpose, the striking distance was finally selected for 

 such measurement. A stout copj^er wire capped with a brass ball at each end 

 was placed inside the gun-barrel : at one end it was brought into metallic con- 

 nection with the gun-barrel by means of tinfoil, but insulated from it for the rest 

 of the distance by a glass tube : an L shaped piece of the same kind of wire was 

 bound to the other end of the gun-barrel by securely wrapping with copper 

 wire ; this L piece was also capped with a brass ball, the balls of the exterior 

 and interior wire being opposite each other, and about three inches apart : a 

 branched rod capped with brass balls, and terminating in a wire which could be 

 brought in contact with the outside of a Leyden jar completed the apparatus, 

 which is figured in cut IV. : 



'O^ 





The arrangement in this apparatus so far as regards external and internal 

 conduction is similar to Prof. Henry's; it differs from his mainly in the fact 

 that the relative amount of conduction is not measured by the action of a mag- 

 netizing spiral, but by the distance across which a spark will leap. If conduc- 

 tion is solely at the surface, and there is no conduction by the interior mass of 

 a conductor, then if the balls of the branched rod are placed at the same dis- 

 tance from the balls of the interior wire and the L wire, when a Leyden jar is 

 discharged through this apparatus from + to — , the spark should always leap 

 from the ball of the L rod to the contiguous ball of the branched rod, and 

 never from the ball of the interior wire to the corresponding ball of the 

 branched rod. When the experiment was tried under these conditions the spark 



