On definite Proportions. s69 
For all other appearances of amalgamation produced by the 
electrical column under similar circumstances. To dispute 
Tespectigg experiments with ammonium or its amalgam is 
a loss of labour; for it is scarcely possible tu devise so de- 
cisive an experiment as not to be explicable on either by- 
pothesis. But whatever is demonstrable respecting the 
fixed alkalis, and the properties which they have in corm- 
mon with ammonia, may also be considered as demon= 
strated of ammonia. The existence of oxygen in the fixed 
alkalis being therefore so indubitably well established, it 
appears that perhaps there is more than simple probability 
in the opinion that ammonia is an oxidated substance. 
When ammoniacal gas is decomposed by electrical shocks, 
we obtain, accarding to the experiments of MM. Henry 
and Berthollet the younger, nitrogen and hydrogen, without 
a trace of oxygen: hence it follows that the oxygen of am- 
monia must be contained in these two gases. One of these 
substances, without doubt the nitrogen, must therefore be 
in a higher degree of oxidation than the other: hence the 
oxygen of the nitrogen must hea multiple by 14, 2,3.... 
of the oxygen of the ammonia: and the true multiplier 
seems to be 14.. The considerations on the different de- 
grees of oxidation of sulphur, which I have already ad- 
vanced, have led me to the conjecture, that every apparent 
multiplication by 14 is in reality a multiplication by 6 or 
12 with regard to some lower degree, which, if it cannot 
exist independently, may at least be found in a combina- 
tion with other substances. If therefore we consider hy- 
drogen as being in this degree, it must contain either 1 or 
+z as much oxygen as nitrogen does; and from the pro= 
portion of the component parts of ammonia, it may easily 
be calculated, that hydrogen can only contain 7g as mucha 
oxygen for every 100 parts of ammonium as nitrogen does. 
Now if in ammonia 100 parts of ammonium are combined 
with 88°2768 of oxygen, in nitrogen they must be com- 
bined with half as much more, or 132°4152, and in hydro= 
gen with one twelfih of this quantity, or 11°0346. Hence 
we should have for Hydrogen 
Ammonium 90*062 100‘0000 
Oxygen.... 9°938 11°0346 
And for Nitrogen 
Ammonium 43°027 10070000 
Oxygen... 56°973 1324152 
Mr. Gay-Lussac attributes to ammonia 18°475 parts of 
hydrogen and 81°525 of nitrogen: according to what we 
have just calculated, $1°525 of nitrogen would contain 46°43 
of oxygen, and 18°475 of hydrogen 1°8 of oxygen, which 
together 
