CjESIUM-OXYGEN-SILVER photoelectric cell 343 



formed by admitting oxygen to a cathode surface covered with a 

 thick adsorbed layer of metallic cajsium. He finds a maximum 

 activity at a cajsium-oxygen weight ratio of 110, which corresponds 

 to 13.2 atoms of caesium per atom of oxygen. This high value may 

 be due in part to an incomplete yield from his caesium pellets, but we 

 doubt that the surface structure so obtained can be closely compared 

 with that which we have described. The cells described elsewhere in 

 his paper have much more closely analogous surfaces. 



N. R. Campbell ^ has performed experiments in which caesium is 

 diffused slowly into a cell containing an oxidized silver cathode, all 

 held at 184° C. in an aniline bath. He finds the completion of the 

 reaction corresponds to the formation of CS2O, or an atomic ratio of 2. 



Kingdon and Thompson ^ also report their cathode surfaces, 

 presumably the same as in the cells described by Koller, to consist 

 of CS2O. 



A Survey of Macroscope Parameters 

 With uncertainty as to the microscopic physical structure of this 

 surface, and no adequate theoretical basis for the correlation of 

 photoelectric response to such a detailed structure, a macroscopic 

 frame of reference has been essential to the correlation of data. This 

 has been particularly important in the development phases of the 

 project so that the results obtained should be self-sufficient and the 

 systematic investigation of substances and processes should not be 

 conditioned by the anticipated interpretations. From a thermo- 

 dynamic standpoint, the surface may be considered as a two-dimen- 

 sional phase or system with two variable components, caesium and 

 oxygen. If this system were in equilibrium, its properties would 

 depend only on its temperature and the amounts of the components. 

 Since it is far from equilibrium the specification of its condition or 

 "state" requires also its past history, the most vital part of which is 

 the heat treatment. 



As the activity of a completed cell is definitely affected by each of 

 the factors controlled in the quantitative technique, it will be seen 

 that the state of the cathode surface is a function of (at least) five 

 parameters, viz: the amount of caesium, the amount of oxygen, the 

 surface roughness of the silver, and the temperature and time of the 

 heat treatment. Some qualitative experience seems to indicate that 

 variation in surface roughness shows itself chiefly by affecting the 

 efficiency of the glow discharge in the quantitative deposition of 

 oxygen, and to a limited extent involves a greater or less extension 



^ Loc. cit. 



* Kingdon and Thompson, Physics, 1, 343 (1931). 



