270 ILLINOIS STATE ACADEMY OF SCIENCE 



neon, helium, argon and carbon dioxide were used. Ni- 

 trogen, neon, air, and carbon dioxide gave about the 

 same degree of signal response, argon gave 25 to 50 per- 

 cent louder response, helium gave a slightly weaker re- 

 sponse, and hydrogen a very weak response. It was 

 found that adjustments of plate voltage and filament cur- 

 rent were critical for all gases to the same degree ex- 

 cept for helium. Helium filled tubes are less critical as 

 regards plate voltage and filament current adjustment. 

 Critical characteristics of plate voltage adjustments for 

 various pressures and gas content are shown in Fig. 2. 



Referring again to Fig. 1 it will be noticed that for 

 very low degrees of vacuum the ''operating voltage" ap- 

 proaches the ionizing potential of the gas in the tube. 

 For nitrogen, hydrogen, or neon, it is about 18 volts, but 

 for helium about 25 to 28 volts, which values represent 

 the values of ionizing potentials for these gases. Ac- 

 cordingly, it should be possible to introduce a gas having 

 a very low ionizing potential, and thus have a tube re- 

 quiring a low operating voltage. Mercury vapor was 

 introduced, and it gave an operating voltage of 14 or 15 

 volts. (Note that the ionizing potential of mercury va- 

 por is 10.5 volts.) A small amount of mercury was put 

 into two tubes, and after being pumped out and sealed 

 off the pressure of mercury vapor pressure could be va- 

 ried by varying the temperature of the tube walls. For 

 best results the mercury vapor pressure was that corres- 

 ponding to a temperature of about 25° C. At tempera- 

 tures of 50° or 60° C the adjustments become very criti- 

 cal ; at 25° C these tubes showed about 50 percent in- 

 crease in audibility over good vacuum tubes. This seemed 

 to indicate that certain metallic vapors would be desir- 

 able if the vapor pressure did not become too great when 

 the tube walls were hot. 



Recently this latter phase of the problem was again 

 taken up, and three of the alkali metals were introduced 

 into detector tubes. The vapors of potassium-sodium 

 alloy, rubidium, and caesium have ionizing potentials of 

 four volts and less. Potassium-sodium alloy in a liquid 

 state similar to mercury was introduced without expos- 



