494 REPORT — 1894. 



in dried oxygen, although the elements of which it is composed do not. 

 Oxygen forms ozone under the influence of the electric discharge as rapidly 

 when dry as when moisture is present. It may be that the substances 

 have not been sufficiently purified, but I believe it may be due to another 

 cause. In many of tliese cases in which water- vapour appears to play no 

 part, we are dealing not with molecules but with atoms. If an atom of 

 oxygen will not unite with another atom of oxygen, then by the decompo- 

 sition of dried potassium chlorate we should perhaps get a gas composed 

 partly of atoms and partly of molecules, and by the decomposition of 

 silver oxide, if its molecular formula is Ag,0, the gas evolved might be 

 composed only of atoms. Iii these cases, however, molecules of oxygen 

 only are obtained. It may be, therefore, that, whatever be the state of 

 dryness of the substance, atoms will always combine. Similarly with re- 

 gard to the combustion of carbon bisulphide, though I used all possible care 

 in its purification, yet it always burnt in dried oxygen. It was noticed, 

 however, that the decomposition point of carbon bisulphide, when heated 

 in a neutral gas like nitrogen, was a little below the point of ignition 

 when heated in oxygen. Therefore, in the latter case, I was dealing not 

 with carbon bisulphide and oxygen but with carbon bisulphide in a decom- 

 posing state, carbon and sulphur being set free not in their ordinary state 

 but in a condition in which they would combine with dried oxygen. This 

 fact is most easily explained by supposing that when carbon bisulphide is 

 heated it splits up into atoms of carbon and sulphur, and that these then 

 combine with oxygen in absence of moisture. 



With regard to the explanation of the effect of moisture on chemical 

 actions in general, several hypotheses have been suggested. The first, that 

 of Professor Dixon, is that the water molecules present undergo an actual 

 decomposition. In the combustion of carbon monoxide, for instance, the 

 «as takes up oxygen from the water, liberating hydrogen, which then 

 combines with the free oxygen, re-forming water. I venture to think that 

 this hypothesis, although I believe it explains all the known facts, is open 

 to one or two objections. For instance, if we accept Berthelot's law of 

 maximum work, there seems to be no reason why water should be decom- 

 posed by red-hot carbon rather than oxygen molecules, since the direct 

 action on the oxygen liberates a fai- greater amount of energy. Dr. 

 Tra.ube has suggested that the explanation is dependent on the oxidation 

 of water rather than on its reduction ; that hydrogen peroxide is first 

 formed by direct union of water-vapour with oxygen, the peroxide again 

 beino- reduced to water, giving up its extra atom of oxygen to the combus- 

 tible. This hypothesis seems to be inadequate in many respects, since 

 many actions in which water plays an important part, e.g., the action of 

 sodium on chlorine, or the combination of ammonia and hydrogen chloride, 

 free oxygen is not present, and, tlierefore, hydrogen peroxide could not be 

 formed. 



Mr. Harcourt fii'st suggested, in 1886, that the explanation of the action 

 was to be sought from a physical, rather than from a chemical, point of 

 view. Dr. Armstrong proposed in the same year a hypothesis, which he 

 calls that of ' reversed electrolysis,' which supposes that no chemical action 

 can take place without the presence of a third body, which must be an 

 electrolyte. With regard to this theory I hope to be able to say more 

 later. I am engaged upon an investigation whose object is to find out 

 what substances can replace water in chemical action, and it may be found 

 to be the case that all such substances are electrolytes. 



I have been engaged during the last two years in an effort to investi- 



