142 DETERMINATION OF NITROGEN. 
If an organic matter, not conlaining nitrogen, be fused with 
hydrate of potassa, the water of the hydrate is decomposed, 
its oxygen unites with the carbon and hydrogen of the organic 
compound, while its own hydrogen is set free in the gaseous 
state. The products formed during this energetic action, vary 
according to the temperature to which the mixture is exposed, 
and according to the composition of the organic compound. 
It is sufficient, however, for us to remark here, that nitrogen 
being absent, hydrogen is liberated in a free state. If the 
substance submitted to this action contain nitrogen, then the 
hydrogen combines with the whole of the nitrogen, and forms 
ammonia. This property, up to the present time, has not 
been taken advantage of, except to detect the presence of ni- 
trogen in any matter. 
With substances containing much nitrogen, such as uric 
acid, melamine, mellone, &c, the whole of the nitrogen is not 
at the commencement of the decomposition used up in the 
formation of ammonia ; part unites with a portion of the car- 
bon of the matter, to produce cyanogen, which in this form 
combines with the metal of the alkali, or as cyanic acid with 
the alkali itself. The fixed nature of the cyanic compounds, 
at a high temperature would lead us to presume, that it might 
be impossible to convert the whole nitrogen into ammonia. 
But the direct experiments which we have made, have shown 
that by the use of a sufficient excess of the hydrate of the al- 
kali, and of a sufficiently high temperature, the whole cyanic 
combination, or compound of nitrogen, if not under the form 
of nitric acid , undergoes such a decomposition, that all the ni- 
trogen is obtained as the final product, under the form of am- 
monia. 
If cyanide of potassium, cyanate of potassa, or para-cyano- 
gen be fused with an excess of hydrate of potassa, at a red 
heat, or if at the same temperature, these bodies be heated 
with a non-fusible mixture of hydrate of potassa or of hydrate 
of soda and quick lime, a considerable amount of ammonia is 
produced, and the residue does not yield the least trace of 
cyanogen or its compounds. It is necessary in this case to 
