EVAPORATION OF ATOMS 



353 



KilHan ^^ and Becker.^ Langmuir and Kingdon ^ have shown that this dis- 

 continuity indicates the existence of caesium in two surface phases. 



Becker ^ was the first to point out that the existence of the two stable 

 phases was a consequence of the relation between Va + Vp and 6. 



Fig. 19 shows both atom and ion evaporation rates (calculated) in the 

 region where v^ and Vp are of comparable nmgnitude. The full line ABC 

 gives the sum of atom and ion rates from a tungsten filament at 848° K. 

 in caesium vapor and exposed to an accelerating field for ions. It is seen 



/.oo -, /.as 

 - /oyr 



Fig. 18. Exponential increase of positive ion evaporation rate followed by discontinuous 



rise to maximum value of Vj,, (vp = ^o). 



that for constant values of [Xa(= v^, -|-Vp) there are three values of 0, as 

 indicated, for example, by the intersections of the dotted line with the curve. 

 Thus at a given pressure two values of 6 may exist corresponding to two 

 phases, a dilute (/?) and a concentrated (a) phase. The intermediate 6 is 

 unstable. These phases are separated by a distinct boundary whose move- 

 ment gives the observed rate of change of Vp at the discontinuity. Migration 

 has been shown to exist at the boundary between the phases and the velocity 

 of propagation has been used to measure D, the coefficient of surface 

 diffusion in a recent note ^^ by the authors. 



The dotted line in Fig. 19 and ABC enclose the two areas Xi and Xo. 



" T. J. Killian, Phys. Rev., 27, 578 (1926). 



*° I. Langmuir and J. B. Taylor, Fhys. Rev., 40. 463 (1932). 



