on the new method for the estimation of Nitrogen, §c. 29% 
In another experiment 20 grms. of metallic tin were melted 
with fresh fused hydrate of soda in a thin U-formed tube, 
with access of atmospheric air, so that during the continuance 
of the experiment a fresh quantity of air was always brought 
into contact with the nascent hydrogen ; I obtained thus 0:008 
ammonio-chloride of platinum = 0°00057 nitrogen. In an 
experiment in which nitrogen was used instead of atmospheric 
air, a similar result was obtained, namely 0°007 ammonio- 
chloride of platinum. 
These experiments prove that the nitrogen of the atmo- 
sphere can in no way form ammonia with hydrogen in a nas- 
cent state. The extremely small quantities obtained in most 
cases must consequently have some other source, which it is 
very difficult to avoid. This however may be attained by the 
following method :— 
Hydrate of soda was melted in a silver crucible until it be- 
came liquid, and then mixed with a small quantity of pure iron, 
reduced from the oxide by means of hydrogen. ‘This was 
readily oxidized with disengagement of hydrogen gas; it was 
then poured into a silver dish previously warmed, and after 
it had cooled was broken into pieces and introduced into a 
slightly curved tube of hard glass half an inch in diameter, 
previously ignited; from 4 to 5 grammes of pure iron re- 
duced from the oxide by hydrogen, were then immediately 
added ; the tube was heated by charcoal placed under it, and 
nitrogen or atmospheric air passed through it. On the first 
passage of the air an extremely small quantity of ammonia 
was generally detected by means of dahlia-paper, or by a rod 
moistened with dilute hydrochloric acid; but this disengage- 
ment of ammonia was only observed for a short time, and 
always ceased before the evolution of the hydrogen, from 
the oxidation of the iron, began. When this took place it 
was connected with the absorption apparatus, and the alkali 
kept in a state of fusion until all the metal was oxidized. By 
carefully following this plan I never obtained ammonio- 
chloride of platinum. 
The same experiment was repeated with a like result with 
perfectly pure crystallized tin, as it is easily obtained when a 
polished rod of tin is suspended in a vessel in which water 
with a little hydrochloric acid, rests on a concentrated solution 
of tin; after one or two days a splendid crystallization forms. 
If the metal happened to be touched by the fingers, or allowed 
to remain exposed to the air before the experiment, a disen- 
gagement of ammonia invariably occurred; but not when it 
as well as the alkaline hydrate were fused just before being 
employed. Pure tin is with great difficulty oxidized by 
U2 
