THE THIRST OF SALTED WATER 647 



does. But has Professor Armstrong forgotten that liquid 

 hydrogen chloride, like pure water, is an insulator for the 

 electric current, as was found long ago by Gore, an observation 

 afterwards confirmed by Bleekrode ? It has been stated by 

 F. Kohlrausch that at ordinary temperatures no pure liquid is 

 a good electrolyte. The theory 01 Arrhenius is still in this 

 point the only one which explains this strange fact ; pure 

 liquids do not conduct, because their molecules have no space to 

 resolve themselves into ions [my italics]. 



It is therefore not improbable that water would conduct 

 electrolytically if we could find a suitable solvent for it. An 

 investigation in this direction would be of very great interest 

 but not without grave difficulties. 



In a postscript to the German edition of the Report of the 

 Leeds discussion (Zeits. phys. Chem. 1891, 7, 418), Ostwald further 

 contended that much greater differences than are apparent 

 between hydrogen chloride and water are noticeable between 

 compounds such as NC1 3 and PC1 3 , N 2 5 and P 2 3 , which are 

 more closely related than are hydrogen chloride and oxide. 



It has always appeared to me that, in quoting such ex- 

 amples, Ostwald made a particularly unhappy choice, as nitro- 

 gen chloride and nitric anhydride are extraordinarily unstable, 

 whilst hydrogen chloride and oxide are both eminently stable. 

 His reference to liquid hydrogen chloride being, like water, 

 an insulator, was also inconsequent. Nothing could show 

 more clearly how little chemical feeling he had at that time and 

 how unable he was to read the facts of chemistry. Moreover, 

 to cite the dissociation of ammonium chloride as in any way 

 confirmatory of the view that hydrogen chloride is dissociated 

 into its ions in aqueous solution is entirely unjustifiable, 

 especially as ammonia is supposed to remain unaffected : the 

 two cases offer no points of analogy. Twenty years ago, it 

 was clear that a sharp distinction must be drawn between 

 the valency of nitrogen in the ammonium compounds and its 

 valency in ammonia — between pentad and triad nitrogen ; all 

 that has happened in the interval serves to confirm this view, 

 especially the considerations advanced by Barlow and Pope in 

 their several memoirs on the correlation of crystalline form with 

 structure. 



It is a striking fact that Helmholtz, notwithstanding his 

 partiality to atomic charges of electricity, was disturbed by the 

 liberties taken by the dissociationists. This is clear from the 



