0\ .NlTROl'S OXIDK IN I'Hl! GAistOL"!^ AND LIQUID STATKS. 221 



there remained a certain quantity, about 7/ij to ^^ of the entire volume, or 

 about 3-5 to 5 per cent. 



This permanent gas cannot be nitric oxide nor oxygen ; for the current of 

 nitrous oxide being made to pass successively through strong solutions of siU- 

 phate of iron and of pyrogallate of potassiiim, these solutions did not change 

 colour. 



The only known permanent gas that could be disengaged is nitrogen. It 

 is a known fact that nitrate of ammonium, in presence of spongy platinum, is 

 decomposed at 160° into nitrogen, nitric acid, and water; the same decom- 

 position of a part of the salt could have been effected by the asperities of the 

 inner surface of the retort. This quantity of nitrogen would exert a con- 

 siderable influence on the specific gravity of the gas. The theoretical specific 

 gravity of pure nitrous oxide is 1-524 ; but being mixed with nitrogen to an 

 amount of 3-5 to 5 per cent., it should be found much smaller, 1-504 to 1-496 

 respectively. This result, however, does not accord with actual experiment. 

 The specific gravity of nitrous oxide, prepared from nitrate of ammonium, 

 Avas determined according to the method of Bunsen (' Gasom. Methoden,' von 

 R. Bunsen) ; for that purpose I used a balloon of 200 cubic centims. Four 

 experiments gave the following results: — 1-531, 1-525, 1-529, and 1-527: 

 the mean value is 1-528, agreeing verj- well with the theoretical specific 

 gravity of pure nitrous oxide, but giving a difference of 0-024 to 0-032 from 

 the specific gravity that would have been found if the gas had been mixed 

 with nitrogen. These differences are too large to be accounted for by ex- 

 jierimental errors. 



An analj-sis of nitrous oxide was made according to a somewhat modified 

 method of Franklaud and Ward. The hydrogen used in these experiments 

 was obtained from the electrolytic decomposition of water, and the oxygen 

 was generated by heating mercuric oxide. To ensure that the mercuric 

 oxide is free from nitrogen, it must be prepared by i)recipitating corrosive 

 sublimate with caustic potash. 



Three analyses of air gave the following satisfactory results : — 



I 



100-00 100-00 100-00 



The following are the results of tlie analysis of nitrous oxide ; — 



I. Nitrous oxide obtained from the liquid 'nitrous oxide of an iron bottle. 



(1) Volume of nitrous oxide used 117-39 



Vohime after the admission of hydrogen 263-62 



Volume after explosion 149-12 



Volume after the admission of oxygen 20 6" 88 



Volume after explosion 160-19 



Hence the volume of the hydrogen 140'23, the volume of the oxygen 

 57-76, and the contraction after the second explosion 46-69. 



The remaining volume (160-19) is a mixture of only nitrogen and oxygen, 

 where the amount of oxygen is 57-76— i x 46-69 = 42-20 ; hence the volume 

 of the remaining nitrogen 160-19—42-20 = 117-99. This vohime is by 0-6 

 larger than the volume of the nitrous oxide used ; hence the amount per 

 cent, is 0-52. 



The amount of hydrogen that remained after the fir^it explosion is | x 



