786 REPORT—1890. 
7. On the Unburned Gases contained in the Flue-gases from Gas Stoves 
and different Burners. By Witu1am THomson, F.R.S.E., F.C.S. 
The author has been working for some time with a view of determining 
whether the gases escaping as flue-gases from gas-stoves and different burners were 
really free from gas capable of combustion, such as carbon monoxide, or unburned 
hydrocarbons or hydrogen, He spent some time in trying to separate and deter- 
mine the quantity of carbon monoxide present, if any; but this problem was beset 
with so many difficulties, that for the moment he abandoned it, and contented him- 
self for the present with determining the quantity of unburned carbon and hydrogen, 
in whatever forms these might exist. For this purpose he arranged an apparatus 
consisting of two carefully-weighed U-tubes filled with strong sulphuric acid, and 
two U-tubes filled with soda-lime, through which the flue-gases were first passed ; 
these absorbed the water and carbon dioxide contained in the flue-gases, leaving 
the hydrocarbons (not absorbed by oil of vitriol), hydrogen, and carbon monoxide 
to pass through a redhot glass tube filled closely, to the extent of 15 inches, with 
oxide of copper prepared im situ from copper-wire gauze. The gases were then 
passed through strong sulphuric acid and soda-lime contained in previously-weighed 
U-tubes, and the results were calculated on the gas measured at 60° Fahrenheit 
and 380 inches barometric pressure, from the measure of gas drawn into the 
aspirator (treated as water-saturated gas), and the carbon dioxide and water- 
vapour absorbed in the tubes were then added on, to make up the measure of flue- 
gas originally employed. 
The coal-gas employed was previously passed through large cylinders filled 
with calcium chloride, to dry it before combustion; and at the time when the 
experiment was going on, and side by side with it, were estimated the carbon 
dioxide and water-vapour present in the air itself, by passing them, by means of 
another aspirator, through strong sulphuric acid and soda-lime in U-tubes previously 
weighed. 
The carbon dioxide in the air itself ranged from 0:41 to 0°66 grains per cubic 
foot of air, and the water from 3°51 to 7:23 grains in the same volume. 
The amounts of carbon dioxide and water collected from the flue-gases, after 
deducting the quantities actually present in the air, were taken as those due to the 
combustion of the coal-gas. 
The standard employed was the one used by Professor Roberts-Austin in his 
analysis of the flue-gases from the burning of coal, and the apparatus employed 
was also generally similar to that employed by him. 
The carbon and hydrogen left unburned were measured in terms of 1,000 parts 
of carbon completely burned, derived from the combustion of the gas in the stove or 
burner. 
The total quantities of carbon dioxide and of water in the flue-gases amounted 
to from 6°6 to 10°8 grains per cubic foot for the former, and from 5’6 to 10°5 grains 
of the latter, in the same volume. 
The amount of flue-gas passed in each experiment was about 1 cubic foot, and 
although the variations were considerable, the general results were conclusive in 
showing that the combustion of gas when burned in gas stoves for heating purposes 
is much more incomplete than one might be led to believe. 
The only burner in which the weights of the tubes remained constant after 
passing the burned gas, and in which the combustion was complete, was in a 
paraffin oil lamp in which the flame was not turned to the highest point. 
Another experiment with the flame turned on gave 12-04 and 3:09 respectively of 
carbon and hydrogen unburned per 1,000 of carbon completely burned. 
The next nearest approach to complete combustion was in an Argand burner, 
in which the carbon compounds were completely burned; but an amount represent- 
ing 0:2575 parts of hydrogen per 1,000 parts of carbon completely burned was 
registered in one experiment, whilst in a second experiment 0:113 of carbon, 
2'5414 of hydrogen were registered per 1,000 parts of carbon completely burned. 
Then came a flat flame, Bray’s burner, burning 4 cubic feet per hour, which gave 
11:12 of carbon and 0°95 hydrogen unburned per 1,000 of carbon completely 
burned. 
