124 Proceedings of the Royal Society of Edinburgh. [Sess. 
happen in this country, are characterised by the almost instantaneous 
discharge into the mine workings of thousands, and in some cases millions, 
of cubic feet of firedamp, and along with the gas hundreds of tons of coal 
are displaced. The coal associated with an outburst is soft, sooty, and 
disintegrated. By taking some anthracite ejected at an outburst in South 
Wales, drying it, filling it into a gas-cylinder, and pumping in pit 
firedamp, it was found that 1 cubic foot of the coal was able to take 
up at 6 or 8 atmospheres pressure, and hold in a condition ready for 
almost instantaneous release, a quantity of firedamp considerably greater 
than could be contained in 1 cubic foot of open space charged with gas 
at the same pressure. The sudden outburst of gas constitutes, in fact, 
a problem in gaseous adsorption ; the gas is held in the coal in a state 
available for discharge when the pressure is released. 
IV. Theoretical Considerations. 
Dr A. M. Williams has shown by an able theoretical analysis * that the 
most probable form for an adsorption isotherm at low concentrations is 
log e g) = A 0 -A ll ; (1) 
where p is the absolute pressure and v the mass of gas adsorbed at that 
pressure by a given mass of the substance. At present it is convenient to 
express v in litres of gas expanded to N.P. and 15° C. A 0 and A 1 are 
coefficients. Williams found the above equation to represent very closely 
the connection between v and p as observed by several previous workers 
who had experimentally determined that connection for various gases at 
pressures below atmospheric. Our own results go to show that the 
equation (with one correcting factor) also holds at high pressure for, at 
any rate, gases above their critical temperatures, though, at those pressures, 
the degree of concentration is considerable. 
To enable this matter to be studied, it was necessary to eliminate the 
effect of the interstitial spaces between the granules. We had previously 
found the interstitial space of cocoanut charcoal (well shaken down) to be 
40 per cent., and that of silica to be 43 per cent, of the gross volume (Briggs, 
loc. cit.), and these values were used in making the correction in question. 
In fig. 3, A is the original curve plotted from the experimental data and 
B is the curve resulting after the volumes compressed into the inter- 
stitial space had been deducted. Now it will be seen that B has a double 
flexure, the gradient first decreasing as pressure rises and then, at the high 
* Proc. Roy. Soc. Edin ., xxxix (1918-19), p. 48 ; Proc. Roy. Soc ., A, xcvi (1919), p. 287. 
