07i the Nature of certain Bodies, 485 



filled with hydrogen gas, and connected with a pneumatic 

 apparatus containing very dry mercury, the quantity of elastic 

 fluid given off, all the corrections being made, equalled thir- 

 teen cubical inches and three quarters, and of these a cubical 

 inch was ammonia j and the residual gas, and the gas in- 

 troduced into the tube being accounted for, it appeared that 

 the elastic fluid generated, destructible by detonation whh 

 oxygen, was to the indestructible elastic fluid, as li'5 to 1. 



In this process, the heat applied approached to the dull 

 red heat. The mercury, in the thermometer, stood at 6^ 

 Fahrenheit, and that in the barometer at 30-3 inches. 



In various experiments on different quantities of the fu- 

 sible substance, in some of which the heat was applied to 

 the tray in the green glass retort, and in others, after it had 

 been introduced into the iron tube ; and in which the tem- 

 perature was sometimes raised slowly and sometimes quickly, 

 the comparative results were' so near these that I have de- 

 tailed, as to render any statement of them superfluous. 



A little more ammonia, and rather a larger proportion of 

 inflammable gas*, were in all instances evolved when the 

 iron tube was used, which I am inclined to attribute to ihe 

 following circumstances. When the tray was brought 

 through the atmosphere to be introduced into the iron tube, 

 the fusible substance absorbed a small quantity of moisture 

 from the air, which is connected with the production of am- 

 monia. And in the process of heating in the retort, the 

 green glass was blackened, and I found that it contained a 

 very small quantity of the oxides of lead and iron, which 

 must have caused the disappearance of a small quantity of 

 hydrogen. 



MM. Gay Lussac and Thenard, it appears from the state- 

 ment, had brought the fusible substance in contact with 

 mercurv, which must have given to it some morsture ; and 

 whenever this is the case, it furnishes by heat variable 

 quantities of ammonia. In one instance, in which I heated 

 the fusible substance from nine grains of potassium, in ft 



* The average of six experiments made in a tube of iron, is ?-4 of inflam- 

 mable pas to 1 of uninflammable. The avera-e of three made in green glas» 

 retorts, is 2-3 to 1. 



H h 3 retort 



