in the presence of dust-free Air and other Gases. 241 



If air, initially saturated and free from all foreign nuclei, be sud- 

 denly allowed to expand, a rain-like condensation results if the ratio 

 of the final to the initial volume, v 2 /vi, exceeds 1*252; no condensa- 

 tion taking place except on the walls of the vessel with smaller 

 expansions. 



When vi\v x exceeds 1252 the condensation continues to be rain- 

 like, showing over quite a wide range no visible increase in the 

 number of the drops w r ith increasing expansion, till a second quite- 

 definite limit is reached when v 2 jvy lies between 1*37 and 1*38. With 

 expansions greater than this the condensation is cloud-like, the num- 

 ber of the drops which are formed increasing at an enormously rapid 

 rate with increasing expansion. If v 2 [vi be made successively greater 

 and greater, a very definite series of colour phenomena is observed as 

 this ratio is increased from 1*38 to 1*44, indicating a very rapid 

 diminution in the size, and corresponding increase in the number of: 

 the cloud particles with increasing supersaturation. Only the smaller 

 apparatus was used in the experiments on the cloud-like condensa- 

 tion, the time taken for the expansion to be completed being much 

 shorter than in the other. 



Experiments were made upon the condensation phenomena in the 

 presence of oxygen, hydrogen, nitrogen, chlorine and carbonic acid. 



Of these, all except hydrogen were alike in showing the two forms 

 of condensation, each requiring a definite degree of supersaturation 

 to produce it. 



Chlorine is too far removed from the condition of a perfect gas to* 

 enable the lowering of temperature and resulting supersaturation to 

 be calculated. This can be done, however, in the case of the other 

 gases. 



By the supersaturation is here meant the ratio of the actual density 

 of the vapour when the expansion has just been completed, and the 

 minimum temperature has therefore been reached, to the density of 

 the vapour in equilibrium, over a flat surface of water at that 

 temperature. 



We may summarise the results obtained with the various gases,, 

 including air, as follows : — 



In order that rain-like condensation may result in the presence of 

 any of these gases, except hydrogen, the expansion must be sufficient 

 to cause the supersaturation to exceed a certain value, amounting, 

 when the final temperature is — 6° C, to between 4 - 2 and 4*4, and 

 diminishing with rising temperature. 



In order that cloud-like condensation may take place in the pre- 

 sence of any of these gases, including hydrogen, the expansion must 

 be sufficient to cause the supersaturation to exceed a certain value,, 

 amounting, when the final temperature is about — 16° C, to 7*9. 



When the supersaturation reached lies between these limits rain- 



