378 PHENOMENA, ATOMS, AND MOLECULES 



With the bulb containing saturated caesium vapor at a temperature Tg, 

 6 does not exceed unity until the filament temperature T is lowered to 

 within 25° of Ts for \i = lo^i (or 80° for |x = lo^s). When T - T^ is 8° 

 (or 24°), 62 is 0.3 so that 30 percent of the surface is covered by a second 

 layer of atoms although 2 percent (or 5 per cent) of the tungsten surface is 

 still bare (due to repulsive forces in this layer) . There are then only enough 

 atoms in the 3rd layer to cover 0.007 o^ the surface. When T — Tg is 0.6° 

 for \i = 10^^ (or 1.9° for \i = 10^^) 62 is 0.9 so that the covering fractions 

 for the successive layers are roughly 0.98, 0.77, 0.33, 0.14, 0.06, 0.026, etc., 

 the total value of o/oi being about 2.3. 



Experimental test of the formation of polyatomic layers 



Fig. 27 shows the building up of a second layer of caesium atoms accord- 

 ing to Eq. (53), for various values of \i. The vertical arrows represent the 

 temperatures corresponding to saturated caesium vapor. 



It was thought desirable to test for this formation of a second layer. 

 At |X(i = 10^^ filament A was maintained, either by passage of small currents 

 or by radiation from filament B, at various temperatures slightly above 

 ^sat. (270°). The temperatures were computed from the measured re- 

 sistance of the filament. 



The adsorbed atom concentration at each temperature was measured 

 by the two filament method. The circles in Fig. 27 give the experimental 

 results. It was found that as indicated by the theory 6 increased only slowly 

 until a temperature within 20° of Tgat. was reached. The increase up to this 

 point was closely that given by Eq. (6) for atom evaporation in the first 

 layer.^^ At v^ 293° a much more rapid rise set in until at 273° a total 6 of 

 1.4 was reached. The direction of the deviation from the theoretical curve 

 shows that atoms evaporate more easily from the 2nd layer of Cs on tung- 

 sten than from metallic caesium. Since 6 depends on the probability of 

 evaporation of the adsorbed atoms in the 2nd layer, application of the 

 Boltzmann equation serves to indicate the amount by which the heat of 

 evaporation differs from that of' liquid Cs. 



«/w' = exp {^.Ve/kT), 



where n and n' are the theoretical and observed concentrations. Since n/n' 

 is approximately 6 between 275° and 295°, AF =:^ 0.045 volt. Thus the 

 theory is extraordinarily well confirmed. There is no tendency to form a 

 second layer until near Tgat.. Atoms adsorbed in the second layer are not 

 held by forces at all approaching in magnitude those holding Cs directly to 



^* This behavior of near i.o makes it likely that Gai is more closely 4.9 X 10^'* 

 than 4.8 X 10^'* as given by the previous less detailed experiments of Section V, 

 introducing a possible error of about 2 percent in the calculations of 0. 



