354 Prof. E. Edlund on the New Electromotive 



current. When in reversing the lever the fork-points n and ri are 

 raised, the surface of the mercury rises about the points, owing 

 to adhesion; and therefore the contact between them lasts longer 

 than if the surface of the mercury remained unchanged. If the 

 current is passing, a formation of sparks ensues as soon as con- 

 tact between the mercury and the fork-points ceases; and as 

 long as the formation of sparks lasts, the current continues. 

 This retardation in the cessation of the current, due to the forma- 

 tion of sparks, increases with the intensity of the current. The 

 reversal of the lever was effected by means of a falling weight, in 

 order that it might always take place with the same speed. The 

 time which elapsed from the moment in which the forked points 

 left the mercury in the cups n and n' to the moment in which the 

 points at the other end of the lever touched the mercury in the 

 cups / and /', was calculated to amount to ^ of a second. Yet, 

 owing to the formation of sparks and the adhesion of the mercury, 

 the time that elapsed between the cessation of the current and 

 junction between / and I 1 was much shorter. That in this rapid 

 reversal the junction between n and n! was not simultaneous 

 with that between / and /', was proved in the following manner. 

 The pole-points at b were pushed together, the conducting-wire 

 p taken away, and in its place a German- silver wire of consider- 

 able resistance was interposed between n f and /'. This resist- 

 ance, however, was so arranged as to produce a smaller diminu- 

 tion in the intensity of the current than the luminous arc when 

 this existed and n' was connected with /' by p. Thereupon 

 the magnetometer-wire g q was detached from q and joined with 

 n'. If, now, the closing had been simultaneous on reversing the 

 lever, part of the principal current must have gone through the 

 magnetometer and have produced a deflection ; in the opposite 

 case the needle must have remained at rest. In this experiment 

 the intensity of the current was greater than in the following, and 

 accordingly the formation of sparks more intense. Yet notwith- 

 standing this, there was no simultaneous closing. In order to 

 diminish in any possible way the formation of sparks, the mer- 

 cury in the cups n and n 1 was covered with a thin layer of oil. 

 That there was no simultaneous closing follows also from another 

 experiment, which will be mentioned afterwards. 



3. We pass now to the actual observations, which, however, 

 are not adduced in the order in which they were made. As the 

 resistance in the luminous arc is great and increases with its length 

 (other things remaining the same), the deflection must have been 

 smaller in proportion as the length of the arc increased. Hence 

 it was important to have the same length of arc in the various 

 experiments. For this purpose one carbon point was moveable, 

 and might be pushed backwards and forwards by means of a 



