290 PHENOMENA, ATOMS, AND MOLECULES 



be expressed in the word 'reflection' is that the atom leaves the surface by 

 a process which is the direct result of the collision of the atom against the 

 surface. 



On the other hand, according to the condensation-evaporation theory, 

 there is no direct connection between the condensation and subsequent 

 evaporation. The chance that a given atom on a surface will evaporate 

 in a given time is not dependent on the length of time that has elapsed 

 since the condensation of that atom. Atoms striking a surface have a certain 

 average 'life' on the surface, depending on the temperature of the surface 

 and the intensity of the forces holding the atom. According to the 're- 

 flection' theory, the life of an atom on the surface is simply the duration of 

 a collision, a time practically independent of temperature and of the 

 mas^nitude of the surface forces. 



To determine definitely which of the two theories corresponds best 

 with the facts, I have repeated Wood's experiments under somewhat 

 modified conditions. A small spherical bulb, together with an appendix 

 containing coconut charcoal, was heated to 600° C. for about four hours 

 while being exhausted by a condensation pump. A liquid-air trap was 

 placed between the pump and the bulb. Some cadmium was purified 

 by distillation and was distilled into the bulb, which was then sealed 

 off from the pump. The cadmium in the bulb was then all distilled into the 

 lower hemisphere. By heating this lower half of the bulb to about 140° C, 

 the upper half remained clear, but by applying a wad of cotton, wet with 

 liquid air, to a portion of the upper hemisphere, a uniform deposit formed 

 within less than a minute and continued to grow, even after the liquid air 

 was removed. 



When the liquid air was applied only long enough to start a deposit, it 

 was found in the first experiments that the deposit did not grow uniformly, 

 but became mottled, or showed concentric rings. The outer edges of the 

 deposit were usually much darker than the central portions. By cooling 

 the coconut charcoal in liquid air, this effect disappeared entirely and the 

 cadmium deposits became remarkably uniform in density. It is thus evident 

 that traces of residual gas may prevent the growth of the deposit, 

 particularly in those places which have been the most effectively cooled. 

 This is probably due to the adsorption of the gas by the cooled metal 

 deposit. This gas is apparently retained by the metal, even after it has 

 warmed up to room temperature, so that vapor condensing on the surface 

 evaporates off again at room temperature. 



These results indicate how enormously sensitive such metal films are 

 to the presence of gas. However, by using liquid air and charcoal con- 

 tinually during the experiments, most of these complicating factors were 

 eliminated. 



