The Mass Spectra of the Alkali Metals. 437 



anode is of necessity the high-potential pole of a discharge- 

 tube arranged to give positive rays, this heating arrangement 

 had to be very carefully insulated. The anode was mounted 

 on a ground joint as indicated so that it could be easily 

 removed and replaced. The discharge-tube was cylindrical, 

 about 4 cm. in diameter, mounted concentric to the axis of 

 the perforated cathode K. A side tube was fitted at B 

 which could be cooled in liquid air; in some of the 

 experiments this was charged with charcoal. 



Hot Anode Discharge Tube. 



_ The anode was placed immediately opposite the perfora- 

 tion of the cathode and about 1 cm. away from it. The 

 platinum strip was bent at one end into a U-shaped channel 

 into which the salts could be melted. The discharge was 

 maintained by a large induction-coil used in the previous 

 work on mass spectra and rectified by means of a valve V. 



Experiments with the Parabola method of Analysis. 



In the preliminary experiments the analysis of the rays 

 was performed by Sir J. J. Thomson's "parabola" method, 

 since this gives the maximum general information, and it 

 was only when suitable conditions and technique had been 

 ascertained that the mass spectrograph was applied. 



The general procedure was to pump out the discharge-tube 

 to the lowest possible pressure, far lower than that necessary 

 to prevent all discharge with the anode cold, and then to 

 heat up the anode until the discharge started. This usually 

 happened at dull red heat, and by very careful adjustment of 

 the temperature and of the primary current in the coil it was 

 possible, under favourable conditions, to maintain a fairly 

 steady current of 1 to 2 milliamperes at a potential of about 

 20,000 volts. 



It ^ will be seen that the arrangement resembles that o( a 

 Coolidge X-ray tube reversed pole for pole, and ir was hoped 

 that it might share the outstanding controllability of that 

 device ; but that expectation was only very partially realized. 



-f- 



