588 Research Staff of the G.E. C, London, on the 



(especially of a tungsten filament) has been studied with 

 great care and with much success by Langmuir *. He finds 

 that if tungsten is heated to a temperature of 2800° K. in 

 gas at any pressure, so that evaporation of the metal is- 

 rapid, all traces of all gases, except the inactive group, can 

 be removed completely and a very high vacuum obtained. 

 He has also shown that in oxygen, nitrogen, and carbon 

 monoxide definite chemical compounds are formed, some- 

 times by the interaction of the gas with the solid metal, but 

 more often with its vapour. Some of the gases can be 

 removed by the tungsten at a much lower temperature - r 

 hydrogen, the extreme case, is removed at a temperature as 

 low as 1300° K. But the removal of hydrogen is a more 

 complex matter. Langmuir interprets the results of his 

 very complex observations as proving that the hot metal 

 converts the hydrogen into an " active " modification, pro- 

 bably with monatomic molecules, which is able to adhere 

 tenaciously to the glass walls or interact chemically in the- 

 cold with oxy o-§n and other elements. 



These actions appear to be purely thermal and determined 

 only by the temperature of the tungsten. When the 

 electric field surrounding the filament is sufficiently great, a 

 discharge occurs which affects the process, usually by causing 

 the disappearance of gas and the progress of chemical 

 reactions which would not take place under the influence of 

 the hot filament alone. Thus in the discharge, nitrogen can 

 be made to combine with tungsten to form the nitride 

 WN 2 with the filament at a temperature much lower than 

 that required to cause the combination in the absence of 

 the discharge. But, in the main, Langmuir'' s observations 

 on the effect of the discharge have been incidental, and his- 

 attention is directed rather to the thermal action. Never- 

 theless, his conclusions are extremely suggestive for any 

 study of the effect of the discharge, and will be referred to- 

 frequently in the subsequent discussion. It appears to be 

 his opinion in general that the nature of the changes which 

 cause the disappearance under the discharge is the same as 

 that of those which cause the disappearance under purely 

 thermal action, and that in both cases the gas disappears 

 because it enters into new chemical relations ; the only 



* I. Langmuir, Journ. Amer. Chem. Soc. xxxiv. p. 1310 (1912) ; xxxv. 

 pp. 105, 931 (1913) ; xxxvii. p. 1139 (1915). It is not always stated 

 in these papers what was the potential difference between the ends 

 of the filament, but since Langmuir recognizes that, if it exceeds 

 40 volts, an electric discharge which affects the disappearance of gas 

 may occur, it may be assumed that the limit was never exceeded. 



