} 1 8 Account of a new Eudiometer. 



which was generated, obtained what he wished. To the 

 air thus produced he gave the name of nitrous air, and, 

 from its possessing the properties of absorbing that portion 

 of atmospheric air which he calls dephlogisticated, first pro- 

 posed its being used as a test for ascertaining the purity of 

 air. His method of proceeding was ingenious and simple ; 

 known quantities of the air to be tried, and of nitrous gas, 

 being mixed, were admitted, after the diminution of volumo 

 occasioned by tlieir union, into a graduated tube, which he 

 denominated a endionietcr. 



It was with the test of nitrous gas that Mr. Cavendish* 

 made his masterly analysis of the air at Kensington and 

 London ; and by many laborious processes and comparative 

 trials, obtained results, the accuracy of which has been more 

 distinctly perceived the more the science of chemistry has 

 advanced. 



The slow combustion of phosphorus, which unites with 

 tlie oxycen to form an acid, and the dccompx)sition of the 

 fluid suiphuret of potash, are certain methods of separating 

 combinations consisting of oxygen and azote; but the de- 

 composition is effected so slowly by tlic action of these sub- 

 stances, that it became a desirable object to discover some 

 means for accelerating the process. This was supposed to 

 have been effected by Guyton, who proposed healing, the 

 suiphuret of potash ; in doing this, sulphurated hydrogen 

 gas, however, is frequently evolved, which, mi-'ving with 

 the residual gas, increases its quantity, and renders the re- 

 sult fallacious. ^ 



The ureen sulphate of iron impregnated with nitrous gasy 

 first discovered by Dr. Priestley, and recently used by Mr. 

 Davy for eudiomctrical purpses, from its possessing the 

 property of absorbing oxygen gas from the atmosphere, i* 

 much t(j^ be preferred to tlie method with nitrous gas, as the 

 green sulphate of iron does not combine with the other ^ascs- 

 with which the nitrous gas is commonly found to be con- 

 taminated, and more certain results are obtained. 



Having had occasion to repeat many of the experiments 



« Phil. Tians. fi>r 1783. 



of 



