384 EXPERIMENTAL INVESTIGATIONS. 



attenuated medium. If the tube be turned one-fourth round 

 its axis in either direction, the discharges reappear, as they also 

 do if the magnets be reversed. The experiment requires care, 

 and it seems to be necessary to make it with a power of coil 

 just sufficient to enable the discharge to traverse the distance 

 between the platinum terminals and with a certain magnetic 

 power : if made with a powerful electro-magnet, the discharge 

 is stopped, whichever side the poles be placed with reference to 

 the platinum wires, but the discharge is only made intermit- 

 tent when the platinum wires are placed in a line joining the 

 poles of the magnet. In the above-mentioned tube the por- 

 tion of one end of the glass around one of the platinum wires 

 was covered with a dark deposit of platinum, from the terminal 

 at this end having been constantly used as the negative pole ; 

 when this was made positive, instead of the clean end, the 

 effect of arresting the current could not be produced : this 

 circumstance, and the delicacy of the conditions necessary to 

 ensure success in this experiment, lead me to think that the 

 stoppage of the discharge is due to the diversion in direction 

 of the electric line of force produced by the magnet ; the dis- 

 charge is, so to speak, blown out by the magnet ; but I hesitate 

 to express a positive opinion on this point. 



When magnets are made to approach the negative pla- 

 tinum there is no stoppage of the discharge, but attractions 

 and repulsions are observed on the blue flame, which have 

 been described by Mr. Gassiot, and which I need not therefore 

 repeat. 



I tried the effect of magnets on the spark from Ruhm- 

 korff's coil in air, but could not arrest them, though they 

 appeared to pass with somewhat greater difficulty and longer 

 intermissions ; the difference was, however, very slight. 



ON SOME ANOMALOUS CASES OF ELECTRICAL 

 DECOMPOSITION. 



Philosophical Magazine, March 1853. 



WOLLASTON showed, In a well-known experiment, * that 

 when Franklinic electricity from the prime conductor of an ordi- 



* Phil. Trans. 1801, p. 431. 



