296 On the new method for estimating of Nitrogen. 
with the disengagment of hydrogen gas at the common 
temperatures, which is by no means the case; it presupposes 
further, that the hydrogen on being set free possesses a far 
greater affinity for the nitrogen than for the oxygen of the at- 
mosphere, which completely contradicts our general experience. 
At high temperatures, where water would be decomposed by 
iron, ammonia is not formed. Kuhlman* obtained only hydro- 
gen and nitrogen, butno ammonia, by the passage of steam, and 
nitrogen over pyrophorous iron heated to a strong red heat. 
I have repeated the doubtful experiment of Austin (at least 
according to the result of Hall+), in such a manner that the 
ammonia contained in the atmosphere (but not its carbonic 
acid) was as perfectly as possible shut out. I introduced into 
a flask of from 4 to 5 litres capacity,some iron nails(one pound) 
previously cleaned from all oxide by dilute hydrochloric acid, 
and then well washed with pure water, and also sufficient 
distilled water to cover the bottom. The flask was connected 
air-tight by an intermediate tube with a second smaller one, 
which contained a small quantity of very dilute muriatic acid. 
By a second hole bored in the cork of the small flask, a tube 
containing asbestus moistened with pure sulphuric acid was at- 
tached, and through which the external air communicated with 
that contained in the greater flask. The object of the hydro- 
chloric acid was to prevent the ammonia formed from passing 
into the sulphuric acid. ‘The air was renewed every day in 
such a manner through a second tube closed with wax in the 
cork of the first flask, that the air entering must pass through 
the sulphuric acid tube. 
After from 14 to 18 days oxidation, the oxide, of which a 
considerable quantity had already formed, was washed out of 
the flask with water and a little dilute hydrochloric acid ; dis- 
solved in hydrochloric acid and the solution to which chlo- 
ride of platinum was added, evaporated nearly to dryness 
in a water-bath. The residue dissolved completely in zether- 
alcohol, and did not deposit a trace of ammonio-chloride of 
platinum even after standing for twelve hours, nor could 
any ammonia be found in the muriatic acid contained in the 
small flask. If the iron was here oxidized at the expense of 
the water, and if the hydrogen by that means set free had at 
the moment of its disengagement formed ammonia with the 
nitrogen of the air, nearly 3 grammes of ammonio-chloride of 
platinum would have been obtained for every gramme of the 
oxide treated in the above manner. ‘This is a quantity which 
could not escape observation. 
* Abhandlung ueber die Saltpeterbildung: Annal. der Chem. und Pharme, 
Bd. xxix, S. 285. 
+ Ann. de Chim. et de Phys, t. ii. p. 42. 
