90 ON THE AMMONIO-IODIDES OF MERCURY. 
mercury with aqua ammonise with the aid of heat, a small pro- 
portion of residue is left undissolved. This I did not examine 
particularly, supposing it to be Hgl, deposited because the 
biniodide did not contain the full proportion of iodine, (which 
is usually the case,) together with a little adherent Hgl 2 . 
According to M. Rammelsberg's observations, it is a com- 
pound analogous in its composition to that which is produced 
when white precipitate is decomposed by washing with water 
or an alkaline solution ; that is, it consists of biniodide of mer- 
cury, oxide of mercury, and amidet of mercury. As the sub- 
ject is interesting from its connection with the theory of ami- 
dogen and the amidets, and, by analogy, with the composition 
of white precipitate, I present an abstract of M. Rammels- 
berg's observations upon it from the Ann. des Mines, not 
having access to the original paper. 
When biniodide of mercury was warmed with aqua am- 
monias, he found a complex action to occur. Besides the com- 
bination of biniodide of mercury and ammonia, which dis- 
solves, a white substance is formed which soon becomes 
brownish, and finally takes the color of kermes, without un- 
dergoing further alteration by prolonged boiling. This brown 
compound does not correspond with white precipitate, but 
with the product which results when the latter is decomposed 
by water or the alkalies, that is to say, it is formed of Hgl 2 , 
HgO, and HgHN 2 . 
In order that a product of this kind should be formed by 
this reaction, hydriodic acid must be produced; and this M. 
R. has found to be the case. 
To obtain this brown compound in a state of purity, the 
supernatent liquid must be rapidly poured off while it is yet 
warm, and the residue boiled again with aqua ammonias, and 
thus repeatedly until no more is dissolved. 
This compound has a brown, sometimes a purple color. It 
is inalterable in the air, and may even be heated to 180° with- 
out decomposition. It does not disengage the least trace of 
ammonia when boiled with a solution of potassa. Judging from 
the results of its analysis, and the nature of the reaction which 
