THE GASES OF THE ATMOSPHERE 127 



of hot iron or in the cold by means of ferrous hydrate. 

 The same difference of density was again observed. 

 Storage of the chemically prepared nitrogen for many 

 months or exposing it to the effect of the silent electric 

 discharge produced no effect on the density of chemical 

 nitrogen and the hypothesis that such nitrogen might 

 be in a condition in which some of its molecules, N 2 , 

 were dissociated into atoms had to be abandoned. 



Early in 1894 the position was therefore as follows : 

 it had been abundantly proved that nitrogen derived 

 from the air by absorption and removal of the other 

 known constituents of air is heavier by about ^^ than 

 nitrogen obtained by the decomposition of chemical 

 compounds. It had also been shown that the apparent 

 lightness of the latter was not due to the presence of 

 lighter impurities nor to dissociation. The only pos- 

 sible hypothesis remaining was that atmospheric nitrogen 

 was mixed with a small quantity of a heavier gas, the 

 nature of which was unknown. , 



About this time Lord Rayleigh's attention was drawn 1 

 to the work of Cavendish described -in his " Experiments 

 on Air " in the Philosophical Transactions for 1785. By 



1 In a lecture on Argon at the Royal Infirmary in 1895 Lord Rayleigh 

 mentions that he derived this suggestion from Professor Dewar. On 

 the other hand, a letter from Eamsay to Lord Eayleigh (28th November, 

 1898), reminds the latter that he had suggested Cavendish " at an early 

 date before 1894 certainly before our conversation at which I asked 

 your permission to look into atmospheric nitrogen." He also possessed a 

 copy of Cavendish's paper with a marginal note, "Look into this," written 

 probably about 1896. The point is not of great importance as to who 

 suggested Cavendish ; that Lord Eayleigh practically attacked the question 

 is the essential fact. 



