﻿the Electric Spark is an Electromotor. 43 



each other so closely that the equilibrium of the needle was con- 

 stant. The position of equilibrium could not, of course, be di- 

 rectly observed, because the needle constantly oscillated ; but a 

 knowledge of it was attained by observing the points on the scale 

 towards which the needle turned. If the positive electricity 

 passed from the comb a to the conduction, the deflection of 

 the magnetic needle was always towards the left of the scale. 

 When now the knobs / and g were pushed so close together that 

 sparks ensued between them, the deflection ought to have been less 

 than before, because part of the current would pass through the 

 conduction cfg e. If the spark at fg were only a conductor of 

 electricity, this must have been the case. But instead of this the 

 deflection was from fifteen to twenty times as much as before. 

 The deflection was in the same direction as if no sparks had been 

 formed between/ and g When the positive current starts from 

 a it divides at c, part flows through the spark at fg, and another 

 part through the galvanometer in the same direction as indicated 

 by the dotted arrows. Now, if there is in the electric spark, as 

 there is in the voltaic arc, a galvanic current in a direction op- 

 posed to that of the discharge, it will traverse the conductors 

 in the direction indicated by the arrows ; hence there must be 

 deflections of the magnetic needle in the same direction as the 

 discharge ; and this was always the case. From this the con- 

 clusion must be drawn that in the electric spark there is an 

 electromotive force which sends a current in an opposite direction 

 to the discharge which the spark causes. 



To test more completely the correctness of this conclusion, the 

 connexion between the spark-apparatus and the conducting wires 

 was altered as follows. This apparatus was removed from the 

 bridge between c and e and inserted at n instead. In the bridge 

 between c and e the rheostat was inserted which had hitherto 

 been at m. As is at once seen, the galvanic current of the spark 

 must now produce a deflection in the magnetic needle in a di- 

 rection the reverse of that of the original current; the experi- 

 ments showed also that this was always the case. The magni- 

 tude of the deflections depends in all these cases mainly on the 

 distance between the knobs /and g ; the resistance in the rheo- 

 stat m, as well as that in h, must therefore be so great that sparks 

 pass between / and g, even if the distance between them is some- 

 what greater. 



3. It seemed also interesting to ascertain whether the voltaic 

 spark could produce chemical decompositions. For this purpose 

 the conductions to the galvanometer were removed, and instead 

 of the rheostat wire h a voltameter was inserted between i and k. 

 The voltameter contained, as usual, dilute sulphuric acid. But 

 a certain proof of the decomposition of water could not thus be 



