a 
s 
ON THE GASES EVOLVED FROM IRON FURNACES. 155 
we possess all the data for estimating the amount of light carburetted hy- 
drogen, carbonic oxide and hydrogen, not only by the formule 1, 2, 3, but 
also by those afterwards described (4, 5, 6, 7) for calculating the quantities 
of light carburetted hydrogen, carbonic oxide, olefiant gas and hydrogen 
contained in a mixture. When both these calculations agree, and when we 
obtain by the last of them e as the value of the olefiant gas, this result may 
be viewed as a certain proof of the complete retention of the olefiant gas and 
other hydrocarbons of unknown composition by the perchloride of antimony 
without any change in the other gases. An experiment instituted for this 
purpose gave the following result :— 
Volume. Sealy i Pressure. 1m. at 0° 
BRMIRCUE LG aca yf uo ods achacgh se asia tions cAsanesess 1554 | —4:3 0:4497 71:00 
After admission Of O.....cccccccscessesescecserecs 343°4 —4:3 0°6351 221-57 
IFLEN COMBUSTION .ccsascerccessacevccccssscucseuss 197°6 —43 0:4872 97°81 
After absorption Of CO. ....scecsecseseeeeceees 122°8 | —3:7 0-4109 51:15 
After admission Of H  ....,cccesecesesceceececees 330°3 | —3'7 06171 206-62 
FA SHED COMPUSHLON -ccccescanasehencvssspaypacceuyens 130-7 | —37 0-4041 53°53 
The relation of aqueous vapour to carbonic acid 02035 grm. : 02113 grm. 
The values deduced for calculation are— 
= 2:3488, 
A = 70°88, 
B= 99°54, 
C = 46°66. 
The formule 1, 2, 3 give us the composition,— 
HA ydrogem. 4. -iissscs sos0dsovsisesee 24022 
Light carburetted hydrogen... 42°73 
Carbonic oxide .....,........00. 3°93 
NiFOMED ssissbieecease ds dacdees ONE 
The formule 4, 5, 6, ‘7 give, on the other hand,— 
FLVQROEON, ..-esnpitan sen eps /savas | 24°H0 
Light carburetted hydrogen... 42:27 
Carbonic oxide,......0..s+se0ee 3°83 
FUMIO tas tc nasned na atnpon- thle 
Olefiant gas ........ s..sseceeeee + 0°28 
The agreement of these results may be considered as a proof of the appli- 
eability of perchloride of antimony for our purposes, as the differences are 
quite within the errors of observation, and as similar differences might arise 
by a variation from unity in the third decimal of the expression Ee But to 
remove every possible doubt as to the accuracy of our results, we have taken 
the specific gravity of the mixture of gases treated with perchloride of anti- 
mony, and compared this result with the theoretical density as calculated from 
the known composition. 
In estimating the specific gravity, it was of importance to operate upon a 
smaller volume of gas than usual, because it was necessary to have the same 
gas collected over mercury, not only in the combustion with oxide of cop- 
per, and in the eudiometric analysis, but also in taking the density of the 
4 
