Hydrogen and Nitrogen by Electron Impacts. 797 



would happen from analogy with hydrogen. When mercury 

 vapour is prevented from reaching the experimental tube by 

 immersing the U-tube in liquid air, a slightly smaller value 

 for "6" is obtained (compare runs S and T). Possibly 

 mercury vapour in the experimental tube helps to "fix" the 

 atomic nitrogen as it is formed. 



It will be noticed (fig. 3) that the value of " b " for 

 nitrogen is considerably smaller than for hydrogen, as long as 

 accelerating potentials below 200 volts are considered. At 

 higher potentials the nitrogen curve crosses the hydrogen 

 curve. 



It is proposed to extend these investigations further. 

 The evidence so far obtained points to the desirability of 

 working at very low pressures, when the simplest conditions 

 are obtained. Possibly a Knudsen gauge or an ionization 

 gauge will be more satisfactory than a McLeod gauge. 



A peculiar clean-up effect. — The first experiments on 

 nitrogen gave very erratic results. These were traced down 

 to an effect which was obtained with nitrogen, but not with 

 hydrogen. If the filament in E were kept warm (i.e., at 

 any temperature above a barely visible red heat) and the 

 tube E surrounded by liquid air, the pressure remained 

 absolutely steady provided that there was no electron current. 

 If the filament heating current were cut off, the pressure 

 fell slowly. In one experiment, the pressure fell from 

 32 x 10 ~ 5 mm. to 4 x 10 ~ 5 mm. in 120 minutes. On heating 

 the filament, the initial pressure was regained in so short a 

 space of time as 30-40 seconds (the liquid air was being 

 around the tube all the time). This cycle of events could 

 be repeated indefinitely, and the initial pressure could 

 always be regained by a short heating. A saturation effect 

 was shown in this type of clean up. Thus, starting with a 

 pressure of 4060 x 10 ~ 5 mm. with liquid air around the tube 

 and the filament at a dull red heat, it fell to 3120 x 10" 5 mm. 

 in five minutes after cutting off the heating current. The 

 fall in the next fifteen minutes was only to 3040 x 10~ 5 mm. 

 On heating the filament for less than a minute, the pressure 

 rose to the initial value 4060 x 10~ 5 mm. The change of 

 pressure due to cooling the tubes by liquid air was only that 

 corresponding to the temperature change. Hence this effect 

 is located at the filament. The only substances in this 

 experiment that can account for this type of clean up are 

 platinum, BaO, and SrO, and possibly the copper leads to the 

 platinum. The effect calls for further investigation. 



Phil. Mag. S. 6. Vol. 41. No. 245. May 1921. :\ (J 



