EVAPORATION OF ATOMS 



351 



Calculations of Vg or Ve/v«, from Eq. (21) give good agreement with 

 the exact Eq. (19) up to of about 0.04. This treatment is probably 

 applicable to dilute films in general. 



In developing an equation of state, Gibbs' equation was given in a 

 wrong form. Correctly it should be dF/d In v^ = nkT as given in the 

 present paper. 



Fig. I in the paper by Langmuir and Kingdon shows electron emission 

 curves plotted as in Fig. 15 of the present paper. The data of Langmuir 

 and Kingdon were not corrected for the cooling effects of the filament 

 leads and were not determined for zero field. Comparison of corresponding 

 points shows emissions in Fig. 15 to be 1/3 to 1/8 of those found by 

 Langmuir and Kingdon. 



IX. POSITIVE ION EVAPORATION (EXPERIMENTAL) 



The rate of ion evaporation was also studied as a function of filament 

 temperature and 6 at various pressures of caesium. The emission varies 



O / 2 , 3 'O. 5 e 



Fig. 16. Current voltage data for positive ion evaporation. 



with field about as much as for electron emission; Vp at zero field was 

 obtained in the same manner as for v,., although (Fig. 16), the break in the 



