Electricity through Argon and Helium. 295 



of r is not important, as the potential-gradient in the Faraday 

 space is small, s is an aluminum wire, used as anode; u is a 

 short length of capillary tubing; and t is a glass bulb into 

 which is sealed another platinum electrode. By using the 

 electrode in t as anode instead of s, a discharge could be sent 

 through the capillary tube u. This was convenient for exa- 

 mining the spectrum of the gas, so as to judge of its purity. 

 An induction-coil was usually used instead of the battery for 

 this purpose, since the discharge obtained with it was much 

 brighter. At v the vacuum-tube was connected with the 

 arrangement previously described for exhausting and admitting 

 the gas. 



B is a telephone, included in the battery-circuit. It is 

 used for ascertaining whether or not the current is continuous. 

 Any intermittence in the current is at once made evident by 

 a singing noise in the telephone, 



For measuring the difference of potential between q and r 

 one of Lord Kelvin's multicellular voltmeters is used. It is 

 represented diagrammatically at C. The range of this instru- 

 ment is from 250 to 1000 volts. 



In order to test the apparatus and gain experience in using 

 it, preliminary measurements were made on the cathode-fall 

 in hydrogen. 



The hydrogen was obtained by the electrolysis of acidulated 

 water ; a test-tube full of it was collected and introduced into 

 the vacuum-tube in the way described for helium and argon. 

 At first the measurements obtained were anomalous, being 

 generally very much too high. It was thought at first that 

 the testing electrode was too far from the cathode. The 

 distance was 9-10 mm. The tube was dismounted, and the 

 distance diminished to its present value 4-5 mm. The 

 readings were still erratic, however. Finally the anomaly 

 was traced to the influence of occluded gases in the platinum 

 wire which formed the cathode. These were got rid of by 

 heating the cathode to a bright-yellow heat for some minutes 

 by a strong discharge from an induction-coil. The pump 

 was then worked, and a fresh supply of hydrogen admitted. 

 Readings were now obtained in good agreement with those 

 of previous observers. The mean of these readings was about 

 298 volts, as against 300 found by Warburg. The agreement 

 is, I think, as good as can be expected. 



Cathode-Fall in Helium. 



The tube was exhausted and helium admitted. At first 

 the spectrum of the gas showed traces of nitrogen; an 



X2 



