112 MR. P. V. BEVAN OX THE COMBINATION OF HYDROGEN 
cases the actions represented by the ordinary equations of chemistry do not occur. 
For example, the actions 
2 CO + Oo = 2 COo, P, + 50o = 2V,0,, 
S + 0 . = SOo, OS 3 + 3 O 3 = COo + 2 SO 0 
do not take place, unless at a very high temperature. But, in the presence of water 
vapour, action goes on readily and the final products are the substances represented 
on the right-hand side of the above equations. Now these actions can all be 
e.\q)lained on similar lines to those already put forward to explain the formation of 
the CfiHjBr molecide. The distinctive character of the explanation is in the 
assumption that an additive complex molecule is formed at first before the final 
products are obtained. In most gaseous actions of this tvpe water vapour is a 
necessary condition of tlie ju’ogress of the reaction. Now there is no molecule which 
shows a greater tendency to form additive compounds than water. This propertv is 
seen in the molecular species winch contain water of crystallization. Bruhl has 
investigated many other “ unsaturated ” properties of water and attributes them 
to tlie quadri-valency of oxygen, but for our purpose all we need to recognise is 
the fact tliat the water molecule tends to form molecular compounds. The part 
played hy the water in a gaseous action we can suppose as follows ; The gaseous 
substance which takes part in the action can unite with the water molecule directly 
forming a molecular compound. This new complex molecule can then exist for some 
time, during which its component atoms can re-arrange themselves, so that an internal 
settling down to equilibrium takes place. In this process a group of atoms may form 
a part of the whole molecule, which has little attraction for the rest of the molecule, 
so that it may split off and give rise to a molecule of the final product. We have in 
effect in this scheme an intermediate compound, but we recognise the possibility of 
this compound being not capable of isolation, and in many cases so unstable that its 
own existence may be practically for an infinitesimally short time. 
Such a theory as the one now put forward requires direct evidence as to the 
existence of the compounds assumed, and, further, that it should lead to expressions 
lor the velocity of the action consistent with the observed phenomena. We shall 
later on discuss the particular case of the hydrogen and chlorine combination ; at 
present we shall examine some of the general results of the theory. 
First of all, to define the kind of actions to which this theory can apply, we may 
divide chemical actions into four classes. We shall not consider actions which o'o on 
witli a large change of energy, as under these circumstances the conditions of the 
action change altogether during the progress of the action, but for our purpose we 
shall suppose the action slow enough for the whole process to be practically 
isothermal. The first class is that in which the actions are simple combinations. For 
example, when an Ag ion combines with a Cl ion and forms an AgCl molecule. The 
