856 Prof. E. Edlund on the New Electromotive 



fork-shaped lever, the current formed a Fig. 2. 



luminous arc at b; when, on the con- 

 trary, the lever was reversed by the 

 falling weight, contact between n and // 

 ceased, the luminous arc was extinguish- 

 ed, and the current passed through the 

 galvanometer so long as the conduction 

 Lasted through the extinguishing lumi- 

 nous arc. The deflection obtained was 

 produced by the electromotive force of 

 the battery, diminished by the electro- 

 motive force in the extinguishing lumi- 

 nous arc. The battery consisted of the 

 same fifty elements as in experiment 1, 

 and with the same resistance in the 

 solution of sulphate of copper. When 

 this resistance was not introduced, the 

 current was so strong that the luminous 



arc. as mentioned before, was not extinguished when the lever 

 was reversed. 



As the mean of ten observations, a deflection of X? 7 * i > divi- 

 sions was obtained, with a probable error o( +1*43, With a 

 luminous an' between silver points no deflection was obtained, 

 the needle was quite stationary. Hence between the silver poles 

 there was no conduction when the luminous are was extinguished. 



Q 



Q 



r 



the deflections obtained ( 1 S 7 and 27*9) the electromotive 



force in the disappearing arc cannot be directly calculated, because 

 the resistance is unequal in the two cases. In the first case it 

 consists of that in the arc, together with that oi' the magneto- 

 meter, In the latter case there is in addition the resistance of 

 the battery and the solution of sulphate of copper. Calling the 

 tirst »/ ami the latter m' , the electromotive force of the battery E2, 

 and the electromotive 1 force in the extinguishing luminous arc ,r, 

 we have 



E-.r x 



in 



■9: 187. 



In the following experiments a determination is obtained oi' 

 the ratio between m! and /;/, by the aid oi' which the calculation 

 can be executed. 



Experiment ;>. The battery consisted of twenty-six elements, 

 ami tin 1 sulphate solution was not interposed in the circuit. 

 When the observations were made as m experiment \, a deflec- 

 tion of 40 divisions was obtained as the mean of fifteen observa- 

 tions, with a probable error of KV5. When afterwards a 

 German-silver wire 18*8 feet long was introduced, a deflection 



