18 Mr. Brande on the composition and analysis 
hydrogen, carbonic oxide, and carbonic acid, presents pecu- 
liar difficulties in the ordinary mode of proceeding ; and as 
it often requires to be performed in investigations relating to 
the gases used for illumination, it became an object to facili- 
tate the process, for which I have used the following plan. 
A hundred measures of the gas are introduced into a gra- 
duated tube, and the carbonic acid absorbed by a solution of 
potassa ; the remaining gas is then transferred to thrice its 
volume of chlorine of known purity, standing over water in a 
tube of about half an inch diameter, and exposed to daylight, 
but carefully excluded from the direct solar rays ; after 
twenty-four hours the carburetted hydrogen and the excess 
of chlorine will have been absorbed, and the remaining gas, 
consisting of carbonic oxide and hydrogen, may be analysed 
by detonation with oxygen in excess ; the measure of car- 
bonic acid formed being the equivalent of that of the original 
carbonic oxide. 
This proceeding depends upon the non-formation of chlo- 
ro-carbonic acid in a mixture of carbonic oxide and chlorine 
in the contact of water, and out of the direct agency 1 of the 
solar rays. Such mixture I have kept several days, occasion- 
ally renewing the chlorine as it became absorbed by the 
water, and have not observed any diminution in the bulk of 
the carbonic oxide. In all these cases it is necessary to as- 
certain the purity of the chlorine by its absorption by water, 
and to be aware of the evolution of common air from water 
during that process. 
7. I repeated many of the above experiments, substituting 
for coal gas a mixture of six volumes of hydrogen with five 
of olefiant gas. The specific gravity of this mixture was. 
