376 Dr. Willows on the Action of a Magnetic 



Crookes * lias shown that the double striae frequently met 

 with in hydrogen are due to the current being carried partly 

 by the hydrogen and partly by mercury vapour. A similar 

 explanation would probably apply to the above results. 



The appearance with hydrogen in the tube was very striking 

 at certain pressures. When the magnet was off, the bright 

 parts of the striae were very narrow and pink in colour, the 

 spaces between being hazy. On putting on the magnet the 

 voltage w r as reduced, the striae became very distinct and 

 steady, double as in air and the interspaces much less hazy. 

 After a few seconds each small stria retreated into the 

 adjoining large one next nearest to the anode. After another 

 interval of a few seconds the voltmeter became very unsteady, 

 the positive column changed from pink to grey, the striae 

 assumed the form they had before the magnet was put on, 

 and the voltage rose to its original value. If the coil was 

 stopped and turned on again after the lapse of 15 seconds or 

 more, the whole of the changes were reproduced ; if turned on 

 at aishorter interval, any particular stage was at once produced 

 according to the time that had elapsed since the stoppage. 



The magnet evidently causes the current to be carried 

 entirely by one gas ; this change occupies several seconds, 

 and the tube must rest for 15 sees, in order to allow the 

 ions of one gas to disappear so that the other may carry its 

 share of current. 



The lower the pressure the smaller the field required to 

 produce double striae. 



I next sought to determine whether the decrease in 

 potential brought about by the transverse field takes place at 

 the cathode, or whether some of this might not be found in 

 the positive column, owing to the increased number of striae 

 found there when the field is on. A tube as shown in fig. 4 

 was used. 



The side electrodes B, C, were thin pointed wires fastened 

 in with sealing-wax joints. B was always in the positive 

 column when A was the anode; at the lower pressures C was 

 in the Faraday dark space or the negative glow. 



The difference of potential between A and the other elec- 

 trodes was measured with the voltmeter both when the 

 magnet was on at D, and when it was off. A large fall in 

 potential between A-B and A-C was always brought about 

 by the field below the critical pressure, the percentage fall 

 over these portions of the tube being much greater than over 

 the whole. By far the greatest potential fall (magnet off) 

 takes place at the cathode, so that the actual fall in volts 



* Proc. Roy. Soc. Ixix. p. 399 (1902). 



