THERMAL DECOMPOSITION OF OZONE 441 



why dry ozone becomes unstable much more rapidly than the 

 moist gas under the influence of the discharge. This difficulty 

 has, however, been quite recently removed through the discovery 

 that in the presence of moisture the oxides of nitrogen and ozone 

 are rapidly converted into nitric acid, which condenses on the 

 surface of the glass and does not to an appreciable extent hasten 

 the decomposition of ozone. 



Before observations on any phenomenon can be used to draw 

 from them conclusions in which any degree of confidence can be 

 placed, it is essential that all the conditions capable of modifying 

 the phenomenon should be known. The velocity of a chemical 

 change can be increased by two causes : firstly, by contact 

 action ; secondly, by the catalytic agency of impurities, diluents 

 and solvents. It ought, therefore, to be determined by experi- 

 ment, in the first place, whether or not the interaction is one 

 that takes place in the gaseous state ; in the second place, what 

 class of catalysts are capable of affecting it. 



In the case of the decomposition of ozone in glass vessels, it 

 is fortunately the case that the glass surface does not to any 

 appreciable extent increase the rate of action, if the surface of the 

 glass be not abnormally large in comparison with the volume 

 content of the vessel. This fact has been established by 

 H. E. Clarke and the writer by the simple expedient of heating 

 ozonised oxygen in vessels of different size and thereby showing 

 that the rate at which the concentration of the ozone diminishes 

 is independent of the surface per unit volume of the enclosure in 

 which the change occurs. This fact, that under the conditions 

 observed the reversion of ozone to oxygen is a case of homo- 

 geneous change, i.e. of change affecting a single uniform state (the 

 gaseous), of course simplifies the interpretation of the results and 

 at the same time imparts to the conclusions a higher degree of 

 trustworthiness ; it enables us to investigate the much-debated 

 problem of catalysis in gases under the most favourable con- 

 ditions and to test the validity of the law of mass action. 



The influence of different classes of impurities on the rate of 

 destruction of ozone has been examined by H. E. Jones and the 

 writer in some detail. The mode of procedure was to heat two 

 samples of ozonised oxygen, one of which contained the impurity 

 under investigation, at the same temperature in two similar glass 

 vessels to which manometers were attached, the experiment 

 being so conducted that the initial concentration of the ozone in 



