114 DELL SYSTEM TECHNICAL JOURNAL 



from matter, and the only known discharges across gases were the 

 discharges in which the gas plays an indispensable role. It is im- 

 portant to note the manner of this evolution, for much of the history 

 of modern physics is dominated by it. We should not be nearly so 

 far advanced as we are, had we not learned two things; how to reduce 

 the amount of gas in a tul)e until an electron can fly clear across it with 

 scarcely any chance of meeting an atom, and how to persuade an elec- 

 tron to emerge from a metal otherwise than by starting a discharge 

 in a gas over its surface. We who are so familiar with the idea of 

 electrons boiling out of a hot wire, or driven out of a cold metal plate 

 by light shining upon it, or fired as projectiles out of exploding atoms, 

 find it difficult to imagine the confusion which of necessity prevailed 

 when all these processes were unknown. In the early stages of 

 research into the discharge in gases, it was made clear that of each 

 self-maintaining discharge a stream of electrons flowing out of the 

 negative electrode is an essential part; the electron-stream maintains 

 the gas-discharge, and reciprocally the gas-discharge maintains the 

 electron -St ream. The latest stage commenced when it was made 

 possible to produce and maintain such an electron-stream inde- 

 pendently of any gas-discharge, and deal with it at will. 



Let me then begin the exposition with this idea, which so many 

 years of research were required to render acceptable: the idea of a 

 stream of electrons emerging from a metal wire or a metal plate, at a 

 constant rate which is not influenced by the presence or absence of 

 gas in the space surrounding the metal. The reader may think 

 either of thermionic electrons flowing spontaneously out of a hot wire, 

 or of photo-electrons fl\ing out of a metal plate upon which ultra- 

 violet light is shining.' 



2. Till. Fi.DW (ir l-"i.i;(TKONs Tiikuk.ii a \i;kv Raricmkd Monatomic 

 Gas, and Tiikir H.ncolnters with the Ato.ms 



Conceive a source of electrons, a negative electrode or cathode, 



which is enclosed in a tube. If the tube is highly evacuated, the 



' While forming one's ideas it is proferalilc to think of the photoelectric source, 

 for a variety of reasons; the electron-stream is not very dense, the electrons emerge 

 with kinetic energies never in excess of a certain sharply-marked limiting value, 

 the metal is cold and not likely to react chemically with whatever gas surrounds it. 

 Also several of the clas.sical fundamental experiments were performed in the years 

 from 180X to I'MV), when the photoelectric elTect had become a reliable instrument 

 of research and the thermionic effect had not. Nowadays it is sometimes used in 

 the ho|>e of >ur|>assinK the accuracy of earlier work, or in experiments on compound 

 (tases which the hot wire might decompose. Still the hoi wire is so much easier 

 to ins<Tt and handle, its emission so much more convenient and controllable, that 

 it will no doubt be cmployc<l in the great majority of experiments in the future as 

 in the past. 



