EVAPORATION OF ATOMS 



339 



straight lines of slope Ba and intercept A. With the values of v^ (= \ia) 

 and i/T at various constant values of a, read from the smooth curves 

 (Fig. 8) drawn through the experimental points, such straight lines were 

 obtained. The intercepts, A, showed a small variation with expressed 

 empirically by 



A = 6i + 4.8(6 - W') . (9) 



With values of A and S from Eqs. (9) and (8) Aa was calculated as a 

 function of 6. From these values of Aa and the observed values of Va and T 

 from the data shown in Fig. 8, Ba was calculated by Eq. (6). For values 

 of up to about 0.6, it was found empirically that Ba for all values of \ia 

 could be represented by 



5„ = 32,380/(1 +0.7145) 



(10) 



within the experimental error of about 0.3 percent. For larger O's, Ba 

 deviated from this expression slightly ; however, the deviation was about 

 2 per cent at (9 = 0.8 and 4 per cent at 5 — 0.9. 



The extent of the agreement of the values of Ba obtained from the 

 experimental data with the empirical equation, Eq. (10), is shown in 



Fig. 9. Experimental variation of i/Ba with 6. Ba is proportional to the heat of 

 evaporation for atoms. X is the deviation caused by the active areas. 



Fig. 9 which is a plot of i/Ba against 6. If the data are to agree with 

 Eq. (10), the points should lie on a straight line. 



In the first analysis of the data by this method, values of i/Ba were 



