256 
SELECTED ARTICLES. 
sorbed in the tube, and the nitrogen and ammonia collected in 
a receiver. By my first experiment, I obtained from 22.21 
of precipitate 0.22 of water; by a second, 0.14 of water from 
20.47 of precipitate. The gaseous mixture, reduced to the de- 
sired temperature and pressure, measured 4.24 cubic inches, 
of which 2.67 were absorbed by water; deducting from the 
remainder 0.23 for the air of the apparatus, there remained 
1.34 per cent., or by weight. 
Ammonia, 0.488 gr. 
Nitrogen, 0.404. 
The following results were obtained by two other experi- 
ments: 
Experiment. Precipitate. Water. Water per cent. 
3 12.14 00.8 0.658 
4 19.42 " " 
Which gives a mean of 0.583 per cent, of water; some of this 
probably arises from the precipitates being inaccurately dried. 
The mean of my analyses gives for the composition of the 
precipitate: 
Mercury, 78.60 
Chlorine, 13.85 
Ammonia, 6.77 
Water, 0.58 
Loss, 0.20 
100.00 
The white precipitate is considered by chemists as a com- 
pound of one atom of peroxide of mercury and one atom of 
sal ammoniac* The great difference in the quantity of mer- 
cury sufficiently proves the incorrectness of this theory, be- 
sides that my experiments have demonstrated that so large a 
quantity of oxygen can not exist in the precipitate. 
* This white precipitate must not be confounded with the white pre- 
cipitate of the Parisian Codex; the latter is protochloride of mercury. 
