8 Prof. E. Edlund's Researches on the Passage 



currents. We are therefore compelled to admit that there must 

 be in the electric spark a proper electromotive force sending 

 a current in the inverse direction of the discharge. As I had 

 before demonstrated that the voltaic arc is the seat of an elec- 

 tromotive force, it was in reality easy to foresee that such a 

 force must also exist in the electric spark. 



Experiment 1. At the two ends of a glass tube of 20 mil- 

 lim. diameter and 60 millim. length, aluminium electrodes 

 were fixed. The distance between the extremities of these 

 wires amounted to 5 millim. A glass tube of the same dimen- 

 sions contained brass electrodes, the extremities of which were 

 furnished with balls of the same metal ; the distance between 

 the nearest points of the latter amounted also to 5 millim. 

 The tubes were connected with the mercury air-pump ; and 

 consequently the air had the same density in both. The dis- 

 tance between the balls e and /(fig. 1) of the electrophorus 

 was 10 millim. At iv a rheostat, presenting a resistance of 

 7*75 ohmads, was placed. This rheostat was composed of a 

 fine German-silver wire wound round four glass columns, each 

 turn of the spiral being sufficiently distant from the neigh- 

 bouring turn to prevent any appreciable induction at the time 

 of the discharge. In this way the following deflections were 

 obtained : — 



Pressure of 



the air, in 



millimetres. 



561-7 . 



353-0 . 



249-0 . 



138-9 . 



72-2 . 



29-5 . 



6-5 . 



4-4 . 



2-0 . 



When the communication between k and the rarefied-air 

 space was interrupted, so that the entire discharge of the elec- 

 trophorus must divide itself between the bridge n and the 

 galvanometer, a deflection of 1'5 scale- division was obtained. 

 Id the experiments cited, a portion of the discharge passed 

 through the rarefied-air space; so that the deflection proceed- 

 ing from the quantity of electricity which came direct from 

 the electrophorus could scarcely amount to more than one 

 division of the scale. If we subtract this from the above-given 



Deflections 



, in scale-divisions, with 





A 



the brass 



the aluminium 



balls. 



poles. 



14-1 



11-8 



9-3 



10-3 



8-8 



7-7 



5-2 



4-6 



3*4 



2-9 



1-7 



3-0 



1-5 



3*8 



3-0 



8-2 



13-3 



16-3 



