370 PHENOMENA, ATOMS, AND MOLECULES 



In the transient experiments at low 6 (Fig. 23) departures from the 

 S.P.P. might be expected to occur. It has been shown that one-half percent 

 of the surface of the filament consisted of "active spots." The final value 

 of 6n for 970° and \ia = 6.2 X 10^^ (curve II) is calculated to be 0.051, 

 but the plotted data show that in about 2 minutes a final value of 0.056 was 

 reached, indicating that the active areas (6 = 0.005) had been completely 

 filled. Now in the transient state, as the filament is slowly coated, the con- 

 centrations in the normal and active areas would remain equal if there were 

 no surface mobility l>y which the active areas could be filled. In 2 minutes 

 the maximum value of ^ (no reevaporation) which could be produced by 

 the arriving atoms is ^— ^o.i6. However, the active areas contribute only 

 1/200 of this or a ^ of v^ 0.0008 and without migration can be filled only 

 after more than 12 minutes. It was particularly observed that the value of 

 (0.056) reached in 2 minutes showed no increase after periods as long 

 as 60 minutes. xA.lso a cold filament nearly completely coated (6^ i.o) 

 when heated to 970° (curve IV) quickly attained the same final value of 

 leaving no doubt as to the absence of any delay in reaching a steady value 

 of 6. The full line branching curves in Fig. 23 calculated with the assump- 

 tion of complete occupation of the active spots by migration early in the 

 coating process, agree excellently with the observed points. The existence 

 of an interphase surface mobility is thus well demonstrated. It is only by 

 virtue of such mobility ^^ that the S.P.P. applies to these experiments with 

 dilute films. 



Since the surface diffusion coefficient for these films is known ^^ from 

 other experiments, it has been possible ^^ to calculate the distance which the 

 adatoms may move in reaching active spots without disturbing the surface 

 phase equilibrium. This distance (v— . 0.03 cm) was found to be large com- 

 pared to the distance (v-^ 0.00 1 cm) between active spots assuming these 

 are located along crystal boundaries. 



XII. MECHANISM OF CONDENSATION AND EVAPORATION 

 FOR CONCENTRATED FILMS 



The fact that a^ = i up to values of 6 as high as ^ = 0.98 is of pro- 

 found significance in its bearing on possible mechanisms of condensation 

 and evaporation. 



According to the reversibility principle ^° "every element in the mecha- 

 nism of a reversible process must itself be reversible," so that "the 

 mechanism of evaporation must be the exact reverse of that of condensa- 

 tion even down to the smallest detail." 



^® See reference 16, discussion of Eq. (12). 



^°I. Langmuir, Jour. Amer. Chcm. Soc, 38, 2221-2295 (1916). See particularly 

 page 2253 and footnote on page 2262. 



