332 Observations on some Comlinalions 



gular exception to the laws of tlie composition of salts ; and as 

 it is rather by analogy than by direct experiments that Gay 

 Lussac was led to adopt this opinion, I have endeavoured to 

 subject this liquid to a rigormis analysis. 



We may effect our purpose l)y placing copper or iron, for in- 

 stance, in contact with the acid in vapour at a red heat. In 

 order to avoid the errors which might result from the action of 

 the acid on the corks employed in the common apparatus, I had 

 at first made use of a glass apparatus, all the parts of which 

 were soldered ; so that the acid could only be in contact with 

 the glass and the metal which ought to compose it : but I met 

 with dilficulties during the experiment, which made mc recur 

 to the employment of a porcelain tube with corks. In conse- 

 quence of some i)recautions which it is needless to mention here, 

 the only error which can result from it is confined to the ab- 

 sorption of a very small quantity of acid by the cork; every 

 thing being so arranged that nothing could be extricated which 

 could complicate the products of the operation. 



The iron or copper in very fine and well polished wires was 

 employed in great excess, and all the oxygen of the acid was 

 absorbed by the metal; the gas not absorbed afterwards ])asscd 

 through a tulie of muriate of lime before proceeding under the 

 bell-glass which terminated the apparatus. It is easy to see 

 that by determining the weight of the metal, that of the tvdje of 

 muriate of lime, before and after the experiment, and the volume 

 of the gas liberated, we might attain the utmost precision. I 

 have . lade by this process several experiments, the results of 

 which have not presented sensible differences. 



In one of those analyses the azotic gas was pure, or at least 

 the quantity of hydrogen wliich it could contain was below the 

 limits prescribed by the usual means for detecting it. In general 

 the proportion of hydrogen was alvvavs extremely small. The 

 following are the details of the experiment in whicii the propor- 

 tion of this gas was the greatest, and in which, nevertheless, it 

 formed only the thirty-two thousandth part of the volume of the 

 azote : 



Weight of the acid analysed . . . . 7'-^35 grs. 



Increase of weight in the iron . . 5fiG0 



Do. of muriate of lime ., .. O'OIJ 



Volume of dry gas at 0" B. O-JG"' . . 1-96 lb. 



100 parts of this gas contain hydrogen . . 3-22 parts. 

 This proportion of hydrogen corresponds to a quantity of water 

 which united to that absori)ed by the nniriate of lime will only 

 farm six thousandth parts of the weight of the acid analysed. 



As this quantity is far inferior to the smallest proportion of 

 ivater which we can admit as the essential element of a com- 

 bination. 



