on the Nature of certain Bodies . 51 
me, that, on the principles of the antiphlogistic theory, it ought 
to be a compound of potassium, a little oxygene and nitro- 
gene, or a combination of a suboxide of potassium and nitro- 
gene ; for the hydrogene disengaged in the operations of 
which it was the result, nearly equalled the whole quantity 
contained in the ammonia employed; and it was easy to ex- 
plain the fact of the reproduction of the ammonia by water, 
on the supposition, that by combination with one portion of 
the oxygene of the water, the oxide of potassium became pot- 
ash, and by combination with another portion and its hydro- 
gene, the nitrogene was converted into volatile alkali. 
With a view to ascertain this point, I made several expe- 
riments on various residuums, procured in the way that I 
have just stated, from the action of equal quantities of potas- 
sium on dry ammonia in platina trays, each portion of metal 
equalling six grains. 
In the first trials, I endeavoured to ascertain the quantity of 
ammonia generated by the action of water upon a residuum, 
by heating it with muriate of lime or potash partially de- 
prived of moisture ; and after several trials, many of which 
failed, I succeeded in obtaining four cubical inches and a half 
of ammonia. In three other cases, where there was reason to 
suspect a small excess of water, the quantities of ammonia 
were three cubical inches and a half, three and eight tenths, 
and four and two tenths. 
These experiments were performed in the iron tube used 
for the former process ; the tray was not withdrawn ; but the 
salt introduced in powder, and the apparatus exhausted as 
before, then filled with hydrogene, and then gently heated 
in a small portable forge. 
H 2 
