Magnetic Field on the Discharge through a Gas. 253 



becoming more concentrated. This could be best followed 

 by producing the appearance gradually by slowly increasing 

 the magnetic field from a very small value. It could then 

 .be seen that positive light gradually appeared from the anode 

 and moved down the tube, while part B of the discharge 

 moved to the side and became redder. Finally the two parts 

 were inseparable, and could not be distinguished by a differ- 

 ence in colour. The discharge in this form was frequently 

 very unstable at the lowest pressures (about *1 mm.) at which 

 it could be produced by the battery-power available, and 

 often ceased after about 30 seconds. During this interval 

 the current was increased by nearly four times its original 

 amount. 



.At a pressure just greater than *1 mm., and for a very 

 feeble current — about 10 galvanometer divisions — the reddish 

 colour at B was absent, as was also the bright part at D. 

 Under these circumstances the magnet decreased the current 

 and caused a contraction of the negative glow if a weak field 

 was used, but if the field was strong, and was suddenly put 

 on at its full value, the discharge ceased altogether. 



Starting with a current in the tube just too small to 

 produce the light at D, a small magnetic field decreased the 

 current ; this decrease was greater and greater as the field 

 was further increased, until when a certain value of the 

 field was reached, the galvanometer started moving back to 

 its original position. Having reached this or when the 

 value of the current was slightly greater than it was at the 

 beginning, the discharge became unstable. D appeared for 

 a few seconds accompanied by a large increase in current 

 and then disappeared again, only to repeat its behaviour. 

 If the field was further increased the discharge became very 

 bright for a second and then ceased. 



If, when the unstable state was reached, the magnetic 

 field was taken off and then suddenly put on at its whole 

 value instead of being gradually brought up to it from 

 smaller values, the discharge stopped at once. When the 

 cathode was a point it was found that the current in the 

 tube increased w T ith the field for all values of the latter that 

 were tried. 



The difference of potential at the terminals altered in the 

 opposite manner to the current ; thus at high pressures the 

 magnetic field increased the potential-difference, while at 

 low pressures the opposite was the case. If the potential- 

 difference be plotted against pressure both when the magnet 

 is on and when it is off, the two curves so obtained cut 

 each other at the same pressure as that at which the curve 



