290 Conduction and Molecular Composition, fyc. [Mar. 25, 



two others, water and sodium hydrogen sulphate, or a new compound 

 is formed, which is easily resolvable into these latter. Moreover, I 

 am inclined to attribute this change and the consequent displacement 

 of the hydrogen in the acid by the sodium, not to the fact that 

 sodium has a greater affinity than hydrogen has for SO4, but to the 

 tendency of hydrogen to displace the sodium in sodium hydroxide and 

 to form water. I do not contend that^in such a case as that quoted 

 direct interaction will take place between the substances as we know 

 them in the solid state ; these may consist of comparatively inert 

 complex aggregates which require to be resolved into simpler molecules 

 either by dissolution or by application of heat. In other words, the 

 presence of water may be necessary, not because it is essential to have 

 an electrolyte present, but because the occurrence of both molecular 

 interaction and electrolytic conduction depends on identical molecular 

 and intermolecular conditions. The chemical interaction takes place 

 entirely independently of the water molecules, and these latter serve 

 only to separate and keep apart the fundamental molecules of which 

 the interacting bodies are composed. 



No final decision for or against the view here put forward can well 

 be arrived at except by the study of the behaviour of gaseous bodies 

 such as ammonia and hydrogen chloride, for example ; if proof can be 

 given that these compounds are capable of directly uniting without the 

 intervention of any third body, a most important step will have been 

 made. 



Other cases deserving of study are the conversion of nitric oxide 

 into nitric peroxide, oxidation by means of ozone, and the action of 

 metals such as sodium on water. As the formation of nitric peroxide 

 involves the prior separation of oxygen-atoms from oxygen-atoms, 

 and not merely the combination of two molecules, it is not improbable 

 that interaction between nitric oxide and oxygen molecules will only 

 take place in the presence of a catalyst. But it is to be borne in mind 

 that both nitric oxide and ozone are bodies which are capable of inter- 

 acting with molecules of their own kind, and that considerable heat is 

 thereby developed ; and it is conceivable that such bodies being pos- 

 sessed of high residual affinity may directly enter into combination 

 with others which have but little residual affinity. As regards the 

 action of sodium on water, the difference in behaviour with dilate 

 sulphuric acid between moderately pure zinc and very nearly pure 

 zinc is so marked that the vigorous action between sodium and water 

 cannot be held to prove much, as no special care is ever taken to pre- 

 pare sodium pure; the question whether the affinity of the oxygen in 

 water for sodium is sufficient to cause their direct association and 

 consequent interaction is an interesting one for experimental inquiry, 

 although it would Le very difficult to make the experiment properly. 



One other application of the theory dwelt on in this paper remains 



