of chlorine and carbon, &c. 51 
and it is evident, that any conveniently formed glass vessel 
will answer the purpose. The admission of water during the 
process prevents the necessity of repeated exhaustion by the 
air-pump, which cannot be done without injury to the latter; 
but to have the full advantage of this part of the process, the 
gases should be as pure as possible, that no atmosphere foreign 
to the experiment may collect in the vessel. 
In order to cleanse the substance, the remaining chlorine 
and muriatic acid were blown out of the vessel by a pair of 
bellows, introduced at the stoppered aperture, and the vessel 
afterwards filled with water, to wash away the muriatic acid 
and other soluble matters. Considerable care is then requisite 
in the farther purification of the chloride. It retains water, 
muriatic acid, and a substance, which I find to be a triple 
compound of chlorine, carbon, and hydrogen, formed from 
the cement of the cap; and as all these contain hydrogen, a 
small quantity of any one remaining with the chloride would, 
in analysis, give erroneous results. Various methods of puri- 
fication may be devised, founded on the properties of the sub- 
stance, but I have found the following the most convenient. 
The substance is to be washed from off the glass, and poured 
with the water into a jar ; a little alcohol will remove the last 
portions which adhere to the glass ; and this when poured 
into the water will precipitate the chloride, and the whole 
will fall to the bottom of the vessel. Then having decanted 
the water, the chloride is to be collected on a filter, and dried 
as much as may be by pressure between folds of bibulous 
paper. It should next be introduced into a glass tube, and 
sublimed by a spirit lamp : the pure substance with water 
will rise at first, but the last portions will be partially decom^ 
