220 THE .ADDRESSES, LECTURES, ETC., OF 



the phonograph, etc., had succeeded in dividing electric currents 

 indefinitely for the distribution of light and power, appears to 

 have taken the public by surprise, and has exerted a most depress- 

 ing influence upon the holders of gas shares. Having given close 

 attention to the question of electric lighting ever since 1867 

 when following the researches of my brother, Dr. Werner Siemens, 

 I presented a paper to the Royal Society describing the dynamo- 

 electric principle I may be allowed to make a few remarks upon 

 the novelty and probable effect of Mr. Edison's startling announce- 

 ment. 



In passing an electric circuit from a main conductor into 

 several or any number of branches, the current divides itself 

 between those branches, according to the well-known law of Ohm, 

 in the exact inverse ratio of the electric resistance presented by 

 each branch. A current may thus be divided, for instance, into 

 ten separate currents of precisely equal force, if each branch is 

 made to consist of a wire of the same length and conductivity ; 

 but if one of these wires was again to be slit into ten wires, 

 presenting in the aggregate the same conductivity, each of these 

 wires would only convey 100th part of the total current. In the 

 same way one of the minor wires might again be subdivided 

 into branches, each of which would convey an amount of electric 

 current which would be accurately expressed by the relative re - 

 sistance of the branch in question, divided by the total resistance 

 of all the branches put together. It would thus seem that 

 nothing could be more easy than to divide a powerful electric 

 current among as many branches of varying relative importance 

 as might be desired ; but in the case of electric lighting a diffi- 

 culty arises in consequence of the varying resistance in each 

 electric light or candle, due to the necessarily somewhat varying 

 distance of the carbon points from each other, upon which the 

 length of the luminous arc depends. In order to work a number 

 of lights upon different branches of the same current, it is necessary 

 to furnish each branch with a regulator so contrived that an 

 increase of current corresponding to too near an approach of the 

 carbon points will produce automatically an increased resistance 

 in that branch circuit, whereas an accidental increase in the 

 distance between the carbon points of any lamp will cause the 

 regulator to reduce the extraneous resistance of the circuit to a 



