on a new detonating Substance. 323 
hands, and should have been severely wounded, had not my 
eyes and face been defended by a plate of glass attached to 
a proper cap, a precaution very necessary in all investiga- 
tions of this body. 
In using smaller quantities and recently distilled mercury, 
I obtained the resulis of the experiments, without any vio- 
lence of action; and though it is probable that some acci- 
dental circumstance might have occasioned the explosion 
of the two grains, yet 1 thought it prudent, in my subse- 
quent experiments, to employ quantities which, in case of 
detonation, would be insufficient to do any serious mis- 
chief. 
In the most accurate experiment that I made, ,2,ths of a 
grain of the compound produced, by its action upon mer- 
cury, 49 grain measures of azote. I collected the white 
powder which had been formed in this and other operations 
of the same kind, and exposed it to heat. [It sublimed un- 
altered, withuut giving off any elastic or fluid matter, which 
there is the greatest reason to believe would not have hap- 
pened, if the compound had contained hydrogen, or oxy- 
gen, or both. The sublimed substance had the properties 
of a mixture of corrosive sublimate and calomel. 
If the results of this experiment be calculated upon, it 
must be concluded that the compound consists of 57 of 
azote to 643 of chlorine in weight, or 19 to 81 in volume; 
but this quantity of azote is probably less than the true pros 
portion, as there must have been some loss froin evapora- 
tion, during the time the compound was transferred, and it 
is possible that a minute quantity of it may have adhered 
to mercury not immediately within the tube. 
The decomposition in this process is very simple, and 
must be supposed to depend merely upon the attraction of 
the mercury for chlorine, in consequence of which the 
azote is set free ; and if the result does not strictly demon- 
strate the proportions of chlorine and azote in the come 
pound, yet it seems at least to show, that these are its only 
constituents. 
As muriate of ammonia and chlorine are the only pro- 
ducts resulting from its action upon solution of muriatie 
acid, it seems reasonable to infer, that this action depends 
on a decomposition of part of the muriatic acid, by the at- 
traction of the new compound for hydrogen to form am- 
monia, which, at the moment of its production, combines 
with another portion of the acid, the chlorine of both com- 
pounds being set free. 
On this view, the quantity of chlorine formed from @ 
x2 certaig 
