As part of a wider investigation into the influence of pressure on the ontaneous ignition of inflammable gas-air media generally, some time go we studied the behaviour of diethyl ether and found it to simulate at of the higher paraffins, in that at low pressures ignition occurs in a high temperature system and at higher pressures in a low temperature system, which develops in the range where normally only cool flames are propagated. We were also impressed with the analogy between our own observations and those on the limits of inflammability of ether-air mixtures made in 1927 by A. G. White who discovered that at low pressures there are two ranges of explosive mixtures which can propagate ame, one for normal and another for cool flames, separated by a range of mixtures through which no flame can be propagated; with increase of pressure these explosive ranges become superposed. We decided to examine the matter more closely because it seemed likely that an explanation of the analogy referred to would throw light on the whole problem. Moreover, the subject is of practical importance, having in mind the risks inherent in certain circumstances in the use of ether as an anaesthetic, and was hoped that our results might also be of some service in this connexion. A likely interpretation could be based on the thermal theory of flame propagation. This applies to the slow initial stages of gaseous explosions and was developed by Mallard and Le Chatelier who proposed the following well-known equation for the velocity, V, of the “uniform movement”:— V = L/C (T -
t
)/(
t
- θ)
f
(T
t
), Where T = the temperature attained in the combustion,
t
= the ignition temperature of the mixture, θ = the initial temperature, L = the thermal conductivity of the unburnt gas, C = its mean heat capacity between θ and
t
, and
f
(T
t
) = a function taking into account the change of L and C with temperature. Other modifications have been proposed, but discussion has usually centred round the above equation; and although it is not possible to apply it quantitatively, owing to the insuperable difficulties involved in determining the precise values of the various terms concerned and lack of knowledge concerning the amount of combustion occurring in the flame front and energy losses, etc., it lends itself well to the qualitative interpretation of the effect of the various factors controlling initial slow flame speeds. For example, Mason and Wheeler showed that with mixtures of like thermal conductivity the speed is proportional to (T —
t
), and inversely proportional to (
t
— θ). Moreover, with combustible-air/oxygen media the mixture giving rise to the maximum flame speed contains an excess of combustible corresponding, owing to the suppression of CO
2
—and H
2
O—dissociation, with that developing the highest temperature; indeed, Bone and Bell have recently shown that with CO-O
2
media the flame speed-composition curve exhibits two maxima corresponding with the suppression of CO
2
dissociation by excess of either CO or O
2
.
Further investigations on the kinetics of gaseous oxidation reactions
