TRANSACTIONS OF SECTION G. 13 
or 
their detailed characteristics are considered in the paper. They confirm the 
general accuracy of our former measurements on cyclical variations of tempera- 
ture in a gas-engine cylinder, and also show some new features due to variations 
from stroke to stroke caused by misfires and the like. 
3. The Lost Pressure in Gaseous Explosions. By Professor W. M. 
Tuornton, D.Sc., D.Eng. 
When the maximum pressure of an explosion is calculated from the heat of 
combustion of the elements of the gaseous mixture values are obtained which 
are in all cases about twice those found by experiment. The mean of a large 
number of ‘ efficiencies of explosion’ for different combustible gases approaches 
one-half. To account for this four chief suggestions have been made: (1) that 
there is dissociation of the products of combustion; (2) that the specific heats are 
much higher at explosion temperatures; (3) that the products are rapidly cooled 
by radiation to the walls of the vessel; (4) that the combustion is not complete 
at the time of reaching the maximum pressure. None of these is in itself sufti- 
cient to account for all the loss of pressure. ‘The suggestion is now made that 
it may be caused by the forces of cohesion which come suddenly into play at 
the moment of formation of a molecule, check the translational energy to which 
alone pressure and temperature are due, and raise for the moment the rotational 
energy of the combining bodies. It is shown that the ratio of the translational 
energy of two colliding and cohering bodies before and after collision is one-half, 
and this ratio is to be expected for the whole mixture. 
The suggestion receives support from the form of the curve connecting 
efficiency of explosion and changed percentage of gas in the mixture. This 
efficiency can be shown to have the form »=1—BN, where B is a constant and 
N is the number of combustible units in unit volume. A combustible unit is 
defined as that group of one molecule of combustible gas and of oxygen atoms 
just sufficient for its complete combustion. At the upper limit N is zero, and 
the efficiency curve is triangular on a base coinciding with the limits of 
inflammability. Its mean height is therefore one-half of the maximum, and this 
agrees very fairly well with the observed values given by Clerk in the case of 
coal-gas. 
4. The Limiting Conditions for the Safe Use of Electricily in Coal 
Mining. By Professor W. M. Tuorntron, D.Sc., D.Eng. 
The paper was a summary of recent researches on the limits of electrical 
ignition of inflammable mine-gases and coal-dust. The lower limit of inflam- 
mability is 5°6 per cent. of methane in air by volume; a temperature of 200° C. 
lowers this to 5:1 per cent. The most inflammable mixtures are at 8 per cent. 
for continuous-current break-sparks, 10°2 per cent. for alternating-current 
breaks. Excess of nitrogen appears to markedly increase the necessary igniting 
current. With non-inductive circuits 1 ampere at 100 continuous volts is a 
typical value ; the corresponding values with alternating current are 7 ampéres 
at 40 periods a second, 16 at 60, 20 at 80, and 29 at 100. By varying the 
inductance the energy of an igniting break-spark is found to be constant at 
about 01 joule. Electric signalling bells have inductance up to 0-5 henry 
and ignite gases at the trembler spark or signalling point. All electric lamps 
and fuses, however small, must be enclosed. Oscillations on a cable sheath 
caused by short circuits on the conductor will not ignite gas, but maintained 
leakage arcs from armouring are only slightly more active than break-sparks. 
Static discharges from 6-inch high-speed belting could not be made to ignite 
gas, nor the blue brush discharge from high-pressure conductors. Movements 
of clouds of dust have been shown to give electrification and to cause sparks, 
but the energy must be much greater than can be obtained experimentally in 
order that this should become dangerous. Wireless telegraphy operations on 
the surface do not induce sparking potentials underground. Capacity sparks 
in general from cables left insulated after being charged are very active, 0°002 
1 Published in the lectrician, \xxiil., p. 822 
1914. LL 
