164 On definite Proportions, 
der, after washing it well, in the sunshine, and then heated it ia 
a small glass retort, I obtained in the receiver some water strongly 
impregnated with ainmonia, amounting to 18°5 per cent. By 
ignition 32 per cent. cf sulphuric acid was expelled, and 49°5 
per cent, of red unmagnetic oxide was jeft behind. It appears 
therefore that, in my earlier experiments, the nitric acid, which 
I had employed for oxidating the iron, and which I then thought 
it unnecessary to mention particularly, had produced a combina- 
tion of a totally different nature from the pure subsalt; and 1 had, 
in all probability, examined a mixture of these substances. This 
ammoniacal combination is very remarkable; it appears to be a 
triple subsalt, analogous to the ammoniacal copper. When heat 
is applied, the sulphuric acid, which had been in combination 
with the ammonia, unites with the oxide of iron, and the am- 
monia is set at liberty. This substance, in its difficult solubility 
in acids, and in its incapability of being altered by caustic alkalis, 
seems to approach in some degree to the triple combinations of 
ammonia with muriatic acid and tin, described by Davy, and to 
the combination with muriatic acid and phosphoric oxide. As 
I had probably obtained only a mixture of this substance with 
the subsalt of the oxide, I thought it useless to attempt a more 
correct determination of its component parts. But I shall en- 
deavour on a future occasion to obtain the combination in a state 
of purity, and to examine the proportions of its constitution. 
{t is demonstrated by the analysis of the pure subsalt of the 
oxide of iron, that the relation between the iron and the sulphur, 
which I had inferred from my first analysis, is incorrect. We 
shall see in the following analyses, that in the subsulphates, the 
oxygen of the acid is either equal to that of the base, or an in- 
tegral submultipie of it. And hence it will follow, that, in all 
the subsulphates, the sulphur is in such a proportion to the metal, 
that its quantity is an integral submultiple of the quantity in the 
sulphuret at a minimum, or ef the quantity in the neutral sul- 
phate of the protexide of iron. In the salt of the oxide of iron 
here described, the proportion of the sulphur to the iron is ex- 
actly one-fourth of the quantity in the sulphuret at a minimum, 
and in the sulphate of the protoxide. 
{ must here call the attention of the reader to a circumstance, 
whiclr is of the highest importance for the completion of the 
doctrine of definite proportions in mixtures, and without attend- 
ing to which; we can scarcely hope ever to see the doctrine of 
the combination of organic bodies sufiiciently illustrated : that is, 
to the existence of the absolute minimum of the combination of 
one sulistance with another, of which all other combinations must 
be multiples. Sinee only very few degrees of combination can 
exist between two bodies separately, that is, without the imter- 
yention 
—— 
