[ 370 ] 



XXXV. Action of a Magnetic Field on the Disckarae through a 

 Gas. By R. S. Willows, M.A., D.Sc> 



WHEX a discharge which is passing through a gas under 

 reduced pressure is acted on by a magnetic field, it 

 is known, both from theory and experiment, that if the 

 field is parallel to the discharge it causes it to pass more 

 easily, while if the field is transverse the opposite is the case. 

 I have shown f, however, that below certain pressures which 

 vary with the conditions of the experiment, a transverse field 

 increases the current in the tube and diminishes the difference 

 of potential at the terminals, provided it be applied near the 

 cathode. At other points of the. discharge a decrease in 

 current is always produced by the field. 



Birkeland has shown J that with the lines of force parallel 

 to the tube, supposed cylindrical, at pressures below *012 mm., 

 the potential-difference is made to diminish by the magnetic 

 field, at first slowly as the intensity is increased, until a 

 certain critical intensity is reached, when a large, abrupt 

 diminution is obtained. 



Almy § has studied these two effects, and has come to the 

 conclusion that they are both due to the same cause. He 

 concludes that the action of the magnet is simply to concen- 

 trate the discharge so that it passes through the gas by a 

 sort of brush or arc rather than in the usual manner, and 

 that this brings about an increase in the conductivity. 



Further study of these effects was my object in starting 

 the experiments described in the following paper. 



When the negative glow is the part of the discharge acted 

 on by a transverse field, an increase in potential at the 

 terminals always takes place ; but it has not been determined 

 previously whether the results depend on whether the 

 magnet acts at the surface of the cathode or at any other 

 point in the dark space. 



In the earlier paper I was fortunate enough to be able to 

 use a large accumulator battery to produce the discharge, and 

 so both voltage at the terminals and current through the tube 

 could be measured. In the present case a coil, driven by a 

 mechanical interrupter, was used, and the voltage only was 

 measured by means of a multicellular voltmeter. Where the 

 range of the voltmeter was not great enough, the tube was 

 shunted by a liquid high resistance, and a fraction of the total 

 voltage taken or the whole measured in steps. An electro- 

 magnet with pole-pieces formed so as to give a very local 



* Communicated by the Physical Society : read January 27, 1905. 



t Phil. Mag. [6] i. p. 250 (1901). 



X Comptes Hendus, cxxvi. p. 586 (1898). 



§ Proc. Camb. Phil. Soc. xi. p. 183 (1901). 



