88 CORRELATION OF PHYSICAL FOECE8. 



action of the gas, or intermedium, through or across which 

 they are transmitted. 



The colour of the electric spark, or of the voltaic arc (i. 

 e. the flame which plays between the terminal points of a 

 powerful voltaic battery), is dependent upon the substance of 

 the metal, subject to certain modifications of the intermedium : 

 thus, the electric spark or arc from zinc is blue ; from silver, 

 green ; from iron, red and scintillating ; precisely the colours 

 afforded by these metals in their ordinary combustion. A 

 portion of the metal is also found to be actually transmitted 

 with every electric or voltaic discharge : in the latter case, 

 indeed, where the quantity of matter acted upon is greater 

 than in the former, the metallic particles emitted by the elec 

 trodes or terminals can be readily collected, tested, or even 

 weighed. It would thus appear that the electrical discharge 

 arises, at least in part, from an actual repulsion and sever 

 ance of the electrified matter itself, which flies off at the points 

 of least resistance. 



A careful examination of the phenomena attending the 

 electric spark or the voltaic arc, which latter is the electric 

 disruptive discharge acting on greater portions of matter, 

 tends to modify considerably our previous idea of the nature 

 of the electric force as a producer of ignition and combustion. 

 The voltaic arc is perhaps, strictly speaking, neither ignition 

 nor combustion. It is not simply ignition ; because the mat 

 ter of the terminals is not merely brought to a state of incan 

 descence, but is physically separated and partially transferred 

 from one electrode to another, much of it being dissipated in 

 a vaporous state. It is not combustion ; for the phenomena 

 will take place independently of atmospheric air, oxygen gas, 

 or any of the bodies usually called supporters of combustion, 

 combustion being in fact chemical union attended with heat 

 and light. In the voltaic arc we may have no chemical union ; 

 for if the experiment be performed in an exhausted receiver, or 

 in nitrogen, the substance forming the electrodes is condensed, 



