AND CHLOEINE UNDER THE INFLUENCE OF LIGHT. 
121 
(4.) The combination is made very much more rapid by the presence of water- 
vapour. 
(5.) Small quantities of impurities joroduce a great decrease in the rate of action. 
(6.) The induction period may be prolonged indefinitely, but its general character 
remains the same. 
(7.) Previous illumination of chlorine, before mixing with hydrogen, makes the 
action, when light falls on the mixture, faster to begin with—shortens the induction 
period. This property of insolated chlorine is lost if the chlorine be bubbled through 
water. Previous illumination of the hydrogen is without efiect. 
(8.) Before any hydrochloric acid is formed, some intermediate body is produced 
which can act as nucleus, on which a cloud can form on a certain expansion. 
II. Theoretical. 
(1.) The induction period is an essential part of the action, and occurs owing to the 
formation of intermediate compounds from water-vapour and the two gases. The 
nature of these compounds is not discovered ; but it seems probable that, first, 
an additive compound of chlorine and water is formed (perhaps Hg = Ov 
then a hydrogen molecule becomes attached to this, forming the compound 
II ^Hg, and finally this complex body breaks down into HgO and 2HC1. 
Cl 
(2.) This view of the process of the combination explains the phenomena of the 
period of induction and the effect of pre-insolation of the chlorine, and makes it 
possible to understand the great effect of impurities. 
(3.) This view of the action of water molecules, in forming additive compounds, 
can be extended to all those actions where water-vapour, as a catalyser, seems to be 
necessary for the progress of the action with finite velocity. 
(4.) In the case of actions in gaseous systems, conditioned by the presence oi a 
catalyser, a period of induction must be expected to occur, and the application of the 
law of mass action to such cases must be made with reference to the intermediate 
compounds formed. There is, therefore, no reason to expect that agreement will be 
found between the theory of mass action as applied to the end pi’oduct equation and 
the actual experimental results. 
In conclusion, I wish to express my thanks to Professor J. J. Thomson for having 
suggested this investigation, and for his advice during the course of the experiments, 
which were performed at the Cavendish Laboratory. 
R 
TED 
C. 
uo, 
1303 
VOL. CCII.-A. 
