270 



Prof. H. E. Armstrong. Electrolytic 



[Mar. 25, 



but lead to one conclusion : that electrolysability is conditioned both 

 by the nature of the elements in the compound and its molecular 

 structure. I have put forward the hypothesis in my address — " that 

 among metallic compounds, only those are electrolytes which contain 

 more than a single atom of metal in their molecules." The mere 

 presence of two or more associated atoms of metal in the molecule, 

 however, probably does not constitute a compound an electrolyte ; 

 and although the hypothesis may be applicable to the majority of 

 simple electrolytes, it certainly does not appear to include all the 

 facts, and it does not serve to explain why certain salts are electro- 

 lytes while others are not. 



The remarkable difference in the electrical behaviour of two com- 

 pounds of the same element, such as stannous chloride, in which the 

 ratio of tin to chlorine atoms is as 1 to 2, and stannic chloride, in 

 which Sn : CI = 1 : 4 — the one being a simple electrolyte ; the other 

 a pseudo- dielectric, if indeed it be not a dielectric — would appear 

 almost to justify the conclusion that in the case of per-salts such as 

 stannic chloride the metal is, as it were, enveloped in a non-conducting 

 sheath of the negative radicle. But whether this be so or not, if — as 

 appears to be the case — all simple electrolytes are metallic compounds, 

 and if only proto-salts are electrolytes, may it not be that electric 

 conduction in simple electrolytes is of the nature of ordinary metallic 

 conduction, differing from it only in the circumstance that the com- 

 pound is decomposed as a consequence of the passage of the current ? 



This would lead to the conception of an electrolyte as being a 

 metallic compound of such elements, and so constituted, that electric 

 conduction may take place through its mass in a manner similar to 

 that in which it takes place through a mass of metal : in fact through 

 the agency of its metallic atoms. On this view, it is essential that 

 the metallic atoms in the molecules comprising a mass of an electro- 

 lyte should be in proximity — as they probably are in proto-salts, but 

 not in many per-salts. The conductivity of two-metal alloys is in 

 many cases much less than that of either of the contained metals : 

 for example, the conductivity of the alloy SnCu 4 is about Jth that 

 of tin and about -^th that of copper. The specific conductivity of 

 metals may, therefore, be much reduced by association with one 

 another; and this being the case, it appears probable that the specific 

 conductivity of a metal would be still more reduced by association 

 with a non-metal, and that if the metal were one of low specific 

 conductivity, it might thus practically become altogether deprived of 

 conducting power : perhaps the " exceptional" behaviour of mercuric 

 and beryllium chlorides is to be explained by considerations such as 

 these. 



To discuss such questions at all satisfactorily, however, we require 

 to know much more of the electrical behaviour of pure fused salts ; 



