114 SIR WILLIAM EAMSAY 



same volume) Ramsay and Cundall proved by ingenious 

 experiments that gaseous nitric peroxide does not com- 

 bine at ordinary temperatures with nitric oxide, also 

 that the density of the gas produced when blue liquid 

 nitrogen trioxide is allowed to evaporate was found to 

 be 22 '35 at ordinary atmospheric conditions. This 

 corresponds to the calculated density of a mixture of 

 N0 2 and N 2 4 with NO and renders the assumption of 

 the presence of N 2 3 highly improbable. 



Ramsay returned to the question three years later. 

 The method of estimating molecular weights by obser- 

 ving the depression of the freezing point introduced by 

 Professor Raoult of Grenoble had recently attracted 

 deservedly increased attention. 



In a paper in the Transactions of the Chemical Society 

 for 1888 (p. 621) Ramsay showed by this method, using 

 acetic acid as the solvent, that nitric peroxide in the 

 liquid state is represented by the formula N 2 4 . The 

 trioxide was found to be unmanageable owing to its 

 instability. But Ramsay took up the question again, 

 and another paper appears in the Transactions for 1890 

 (p. 590), in which a fuller account is given of both these 

 oxides of nitrogen, and the probability that the trioxide 

 in the liquid state is correctly expressed by N 2 3 is con- 

 verted into certainty. It appears to undergo dissocia- 

 tion to some extent even at 90. With regard to a 

 mixture of the oxides of nitrogen in the gaseous state it 

 seemb to have been shown by Dixon and Peterkin some 

 years later that a small quantity of N 2 O 3 may exist in 



