278 



Prof. H. E. Armstrong. Electrolytic 



[Mar. 25, 



aggregates of S0 2 molecules, and these may be to a large extent con- 

 served in concentrated aqueous solutions. But the main explanation 

 of the variation in conductivity on dilution must be looked for, I 

 think, in the peculiar relation which sulphur dioxide manifests to 

 water ; it is more than probable that the initial interaction involves 

 the formation of a hydrate, (S02)x(OH 2 )? / , and that from this on dilu- 

 tion is formed sulphurous acid, SO(OH) 2 , and perhaps also " sulphonic 

 acid," HS0 3 H. Taking into account the properties of selenious oxide, 

 Ostwald's results appear to me in this case again to lead to but the 

 one conclusion, that conductivity increases in consequence of the 

 specific influence of the fundamental molecule of the compound 

 making itself more and more felt as by dilution it becomes more 

 and more disentangled from its fellows. 



The behaviour of solutions of neutral metallic salts on dilution is 

 very similar to that of acids; abundant proof of this is afforded 

 especially by F. Kohlrausch's refined measurements, of which an 

 account has recently been published (" Wied. Ann.," 1886, 26, p. 162). 

 I venture to think that a similar explanation to that above given for 

 oxides will apply to salts ; and also that the low molecular con- 

 ductivities of salts as compared with corresponding acids may be 

 regarded as confirmatory of my hypothesis. I think we must admit 

 that the metals generally have less affinity than hydrogen for negative 

 radicles ; if this be granted, we have at once an explanation of the 

 fact that metallic salts are mostly fixed solids, few of which are more 

 than moderately soluble in water while many are very difficultly 

 soluble or insoluble, whereas the corresponding acids are mostly 

 volatile and readily soluble in water, if not miscible with it in all 

 proportions. The affinity of the negative radicles being less exhausted 

 by union with metals than with hydrogen, the fundamental molecules 

 of salts are more prone to unite together to form complex aggregates. 



Arrhenius, who has studied the electrical behaviour of solutions 

 of a number of salts,* attributes the change observed in molecular 

 conductivity on dilution — as I have done — to molecular changes ; but 

 his deductions are all based on the acceptance of the Williamson- 

 Clausius hypothesis of dissociation. 



My hypothesis would also account for the increase in conductivity 



* " Bihang till Kongl. Svenska Vetenskaps-Akademiens Handlingar." Attonde 

 Bandet. Hafte 2. Stockholm, 1884. Arrhenius, S. : " Recherches sur la Conduc- 

 tibilite Gralvanique des Electrolytes. I. La Conductibilite Galvanique des Solutions 

 Aqueuses extreniement diluees, deterrninee au moyen des Depolarisateurs." 63 pp. 

 II. " Theorie Chimique des Electrolytes." 89 pp. Although aware of his work 

 from Ostwald's reference to it, I was unable to study his memoir until after this 

 paper had been elaborated. Ostwald's quotations, moreover, did not enable me go 

 realise the importance which Arrhenius attaches to the occurrence of molecidar 

 simplification and changes in composition on dilution. 



