CONDENSATION AND EVAPORATION 293 



the atoms striking such a surface will fall in positions adjacent to those 

 atoms already on the surface. With cadmium vapor at 170°, 1.4 X lO^'"' 

 atoms per square centimeter strike the surface each second, so that 

 2.8 X 10^^ would condense in the first second around the 4 X lo'^^ atoms 

 remaining on the surface. Thus in only a few seconds the whole surface 

 becomes covered with a layer of cadmium atoms. This explains why a 

 surface only partially covered with cadmium atoms can serve so effectively 

 as a nucleus. If a much smaller fraction than 0.025 of the surface is covered, 

 however, there is a long delay in completing the first layer of atoms, so that 

 the visible deposit is formed much more slowly. 



The above experiments prove that the range of atomic forces is very 

 small and that they act only between atoms practically in contact with each 

 other. Thus a surface covered by a single layer of cadmium atoms behaves, 

 as far as condensation and evaporation are concerned, like a surface of 

 massive cadmium. This absence of transition layer is in accord with my 

 theory of heterogeneous reactions (10). 



One of the best proofs of the correctness of the condensation-evapora- 

 tion theory was obtained in experiments in which nuclei formed at liquid 

 air temperature, were not allowed to warm up to room temperature, but 

 only to —40° C. In this case the nuclei were formed in one minute from 

 cadmium vapor at 54° C. The nuclei which were kept at — 40° C. developed 

 rapidly into cadmium mirrors in cadmium vapor at 170°, while those at 

 room temperature developed extremely slowly. A still more striking 

 demonstration of the theory was obtained when one of the nuclei was 

 allowed to warm up to room temperature and then cooled to —40° before 

 exposure to cadmium vapor at 170°. This nucleus did not develop nearly 

 as rapidly as that which had not been allowed to warm up to room tem- 

 perature. 



These experiments prove that single cadmium atoms actually evaporate 

 off of a glass surface at temperatures below room temperature, although 

 they do not do so at an appreciable rate from a cadmium surface. 



This theory affords a very satisfactory explanation of Moser's breath 

 figures on glass and the peculiar effects observed in the formation of frost 

 crystals on window panes. In fact, the theory appears capable of extension 

 to the whole subject of nucleus formation, including, for example, the 

 crystallization of supercooled liquids. 



The final paper will be submitted to the Physical Review for publication. 



BIBLIOGRAPHY 



(i) Langmuir, I., Physic. Rev., Ithaca, N. Y. (Ser. 2), 1913, 2 (329-342) ; Physik. 

 Zs., Leipzig, 14, 1913 (1273) ; Langmuir and Mackay, Physic. Rev., Ithaca, 

 N. Y. (Ser. 2), 4,1914 (377-386). 



