Field on the Discharge through a Gas. 



6i < 



between A-C brought about by the magnet was not sufficient 

 to account tor the whole decrease for the tube. When the 

 magnet caused the potential between A-D to increase, it 

 frequently effected a fall between A-B or A-C if the gas- 

 pressure was not above about 2 mms. 



The results given by this tube are. however, not altogether 

 trustworthy, for the sealing-in o£ the electrodes B, C caused 

 the magnet to influence the tube as a whole at a much lower 

 pressure than previously ; that is to say, the positive column 

 was striated for some time before the magnet caused the 



Fijy. 



+ VE 



5 



P" 



B 



voltage to drop. This may be due to the distortion of the 

 discharge by the wires, but it is much more likely on account 

 of B and C acting as secondary anodes and cathodes according 

 to which is connected with the voltmeter. It is certain that 

 a dark space was developed round them in some instance-. 



A tube was next tried in which the development of the 

 positive column could be controlled while everything else 

 remained the same. This is shown in fig. 5. 



A was a pointed aluminium wire and was always made the 



