exposed to Entladungsstrahlen. 77 



ionization of a gas would give rise to pulses similar in character 

 to those produced by the stoppage of an electrified particle. As 

 the stoppage of the electrified particles which constitute the 

 cathode rays is probably more abrupt than the ionization or re- 

 combination of ions in a gas, we should expect the pulses produced 

 by the cathode rays to be thinner and more intense than those 

 which constitute the Entladungsstrahlen. 



The Entladungsstrahlen are absorbed very rapidly by air, and 

 in order to detect any leakage produced by them at atmospheric 

 pressure, I have found it necessary to deal with regions within 

 a short distance of the arc. The arrangement used for experi- 

 ments at atmospheric pressure is shown in Fig. 2. The cylinders 



Eabth 



Fig. 2. 



between which the leak was taken were placed close behind a 

 piece of fine wire-gauze connected to earth ; the gauze was used 

 to prevent any ions in the part of the gas traversed by the electric 

 discharge being dragged between the cylinders by the electric 

 field. The arc from the transformer was about 1 cm. in front of 

 the gauze and a strong air blast was blown between the arc and 

 the gauze to prevent the gas from the arc reaching the cylinders 

 by convection. Under these circumstances when the arc was 

 passing there was a leak between the inner and outer cylinders 

 when these were at different electric potentials. 



An induction coil was then tried instead of the transformer 

 and the leak was again obtained. Using the same length of arc 

 the leak was greater when thin electrodes were used, which got 

 nearly white-hot, than when the electrodes were thicker. This 

 was not due directly to radiation arising from the high temperature 

 to which the terminals were raised ; for there was no leak when 

 the arc was replaced by a platinum wire raised to a temperature 

 which, as far as the eye could judge, was much higher than the 

 temperature of the electrodes used for the arc. There was also 

 no leak when an incandescent gas mantle was substituted for the 

 arc, nor could I detect any from the oxy-hydrogen flame. 



