THE ELECTRIC LIGHT. 423 



oak. I send the self-same current through a wire 

 composed of alternate lengths of silver and platinum. 

 The silver offers little resistance, the platinum offers 

 much. The consequence is that the platinum is raised 

 to a white heat, while the silver is not visibly warmed. 

 The same holds good with regard to the carbon ter- 

 minals employed for the production of the electric 

 light. The interval between them offers a powerful 

 resistance to the passage of the current, and it is by 

 the gathering up of the force necessary to burst across 

 this interval that the voltaic current is able to throw 

 the carbon into that state of violent intestine commo- 

 tion which we call heat, and to which its effulgence is 

 due. The smallest interval of air usually suffices to 

 stop the current. But when the carbon points are first 

 brought together and then separated, there occurs be- 

 tween them a discharge of incandescent matter which 

 carries, or may carry, the current over a considerable 

 space. The light comes almost wholly from the in- 

 candescent carbons. The space between them is filled 

 with a blue flame which, being usually bent by the 

 earth's magnetism, receives the name of the Voltaic 

 Arc. 1 



1 The part played by resistance is strikingly illustrated by the 

 deportment of silver and thallium when mixed together and volati- 

 lised in the arc. The current first selects as' its carrier the most 

 volatile metal, which in this case is thallium. While it continues 

 abundant, the passage of the current is so free the resistance to it 

 is so small that the heat generated is incompetent to volatilise 

 the silver. As the thallium disappears the current is forced to con- 

 centrate its power ; it presses the silver into its service, and finally 

 fills the space between the carbons with a vapour which, as long as 

 the necessary resistance is absent, it is incompetent to produce. I 

 have on a former occasion drawn attention to a danger which besets 

 the spectroscopist when operating upon a mixture of constituents 

 volatile in different degrees. When, in 1 872, I first observed the 

 effect here described, had I not known that silver was present, I 

 should have inferred its absence. 



