oe Cav 
SECTIONAL TRANSACTIONS.—B. 303 
Friday, August 28. 
5, Joint Discussion with Section G on The Ignition of Gases. 
Prof. H. B. Dixon, F.R.S8. 
During the past three years experiments have been made for the Safety in Mines 
Research Board on the ignition of gases. 
I. Adiabatic Compression.—Mixtures of methane and air, and of hydrogen and 
oxygen, have been fired by rapid compression in steel cylinders by a piston which is 
stopped and held rigidly at any desired position in the cylinder. By trial there is 
found a compression which will just fire the mixture. With a rapid self-heating mixture 
(e.g. electrolytic gas) the clamping of the piston at the end of its stroke has a small 
but measurable effect ; with a slow self-heating mixture (e.g. methane and air) the 
clamping of the piston makes an important difference, since the gas during the period 
of self-heating has time to do work on the piston. 
When oxygen is added to electrolytic gas, the ignition-point, as calculated from the 
compression, is continuously lowered; with methane-air mixtures the ignition-point 
is slightly reduced as the air is increased from 90°5 per cent. (the volume for complete 
combustion) to 93 per cent., after which the ignition-point rapidly rises. 
Il. Concentric Tube Experiments.—When a stream of inflammable gas from a 
narrow central tube is made to enter an atmosphere of air or oxygen in a large cylinder, 
both gases being heated equally before meeting, the gas will ignite very rapidly (under 
15 seconds) on escaping from the orifice so long as the temperature is maintained above 
a crucial point. On allowing the furnace to cool sufficiently, an appreciable time 
interval occurs between the turning on of the gas and its ignition. This gradually 
increasing ‘ lag ’ is noted until the gas fails to ignite in 10 or in 15 seconds. The cylinder 
is then slowly heated, and a series of diminishing lags is recorded, until the gas ignites 
again in less than -5 seconds. The means of the falling and rising temperatures observed 
at corresponding lags are taken as single readings. 
A steel case has been made to contain the electric furnace and silica cylinder, so that 
ignitions may be observed both under increased and diminished pressure. An increase 
in pressure, as a rule, lowers the ignition-point of gases as indicated by theory ; but 
in many cases, notably with hydrogen, a diminution in pressure down to one-tenth 
of an atmosphere causes a continuous lowering of the rapid ignition-point. 
IIT. Incipient Combustion.—Most gases show some visible signs of partial com- 
bustion before normal inflammation—notably methane. But the only gases with which 
we have been able to stabilise these incipient flames are mixtures containing carbon 
disulphide or ether. 
Prof. W. T. Davin. 
At Leeds we are working upon an apparatus in which spontaneous ignition tem- 
peratures of inflammable gaseous mixtures are determined by means of adiabatic 
compression produced by the flow of compressed air into a tube in which the 
_ inflammable mixtures are contained. Preliminary experiments with this apparatus 
show that the spontaneous ignition temperature varies with the pressure. For a 
26-5 per cent. mixture of methane and oxygen, the following results were obtained :— 
Initial pressure of mixture (atmo- 
sphere) Ee - 2 si 3 
Minimum final pressure at which 
ignition took place (atmospheres) . 10-25 13-2 15-75 21-0 
Ignition temperature (°C) . 2 Z 475 400 370 355 
Thus, in mixtures of methane and oxygen, the ignition temperature is decreased 
as the pressure is increased. 
Experiments are being made upon the ignition temperatures of inflammable 
mixtures diluted with nitrogen and other gases, but they have not yet progressed 
sufficiently to enable me to give results. 
It is possible that the experiments with nitrogen may yield results of great interest 
in view of experiments whieh, in conjunction with Messrs. S. G. Richardson and 
25 *BY cr MBys {10 
_ W.’Davies, I have been making upon the effect of radiation on the rate of combustion 
