ON ELECTROLYSIS IN ITS PHYSICAL AND CHEMICAL BEARINGS. oOD 



urged against the dissociation electrolytic hypotbesia are unsatisfactory, and that 

 the arguments in favour of the residual affinity hypothesis need a more convincing 

 corroboration than hitherto given. On the other hand, the dissociation electrolytic 

 hypothesis has latterly acquired a highly increased probahility, as the lowering of 

 the freezing-point on addition of salts to water, and the additive properties of salt 

 solutions can be explained by it. In a preliminary notice O&twald has announced 

 that the conductivity of acids and bases can be calculated upon the basis of this 

 hypothesis ; and in a paper soon to be published in the ' Zeitschrift f. Phys. Ch.' 

 I will attempt to prove that the known facts regarding conductivity of mixtures, 

 velocity of interaction, and double decomposition between electrolytes can easily be 

 accounted for on the same hj'pothesis. 



Note on tlie foregoing Rephj. By Henry E. Armstrong. 



It is not, I think, desirable for me to attempt to reply at length to Dr. Arr- 

 henius ; evidently he retains his original position, but apparently he has not 

 appreciated the character of all my arguments. Thus I cannot imagine how he 

 conceives (§ 9) that any conclusion as to the triple union in the acid 

 C,H,.C=C.0O,II being less firm than the simple union in C,H,.CH„ - CH2.C0,H 

 — which I entirely dispute {cf. ' Phd. Mag.' 1887, 23, 103) — affects the question 

 at issue, which is whether phenylpropiolic or cinnamic acid is the more readily 

 separable into its ions, i.e., C^H^.C^CCOo and H in the one case, and 

 CeH,.CIL.CH,.COo and H in the other. 



K in the formula C = E|R (§ 10) was never regarded by me as simply the 

 resistance of the solution in which the interaction takes place ; it is the resistance 

 of the circuit in which interaction occurs. This latter may be quite different from 

 the former, as of a number of substances in solution some only may be capable of 

 entering into the true circuit of change. The influence of added neutral salt may 

 be of a very complex character, and until we know more of the constitution of 

 solutions it is very difficult to arrive at any decision as to its exact character. 



No doubt I exaggerated, and should have said (§ 3) that the resistance of 

 alcoholic solutions is very great in comparison with that of corresponding aqueous 

 solutions. I cannot accept either Fitzpatrick's or Hartwig's results as final, 

 interesting as they are, as they were obtained with ' commercial ' alcohols ; it is 

 very important that ^jwre alcoholic solutions should be examined, but this would 

 be a task of exceptional difficulty. 



Dr. Arrhenius says (§ 7) that diminution in conductivity on addition of 

 alcohol must chiefly depend on the increase of friction with the quantity of alcohol, 

 which is a priori vert/ probable. But is it ? We know that alcohol does occasion- 

 ally form compounds corresponding to hydrates, but as a rule they are far less 

 stable than the hydrates. To the chemist it would therefore seem probable, 

 it priori, that friction would diminish with the quantity of alcohol. Cane sugar, 

 as Dr. Arrhenius points out, appears not to be a conductor in aqueous solution, 

 nor is water: therefore, neither is appreciably dissociated in aqueous solution, 

 according to the dissociation hypothesis ; yet cane sugar may be hydrolysed by 

 mere heating with water alone. This and a multitude of similar facts appear to 

 me to weigh heavily against the dissociation hypothesis.^ 



I cannot believe that the sudden increase in conductivity of solid silver 

 iodide (§§ 1, 2) at about 160° is due to a sudden dissociation of the molecules into 



■ The hydrolysis of cane sugar under such circumstances is conceivably condi- 

 tioned by the presence of metallic salts, derived from the glass vessel in which the 

 operation is performed. A more effective argument is supplied by the extraordinarily 

 rapid hydrolysis of cane sugar, which is induced by the unorganised ferment inveHase, 

 derivable from yeast. As typical of the mode of action of such a substance, I may 

 refer to the hydrolysis of the isocyanides R.NC of Gautier and Hofmann, which are 

 unafEected by alkalies, but immediately attacked by aqueous acids : it is .significant 

 that these cyanides form crystalline componnds with hydrogen chloride and a number 

 of anhydrous acids, and that the compounds at once interact with water. 



A A 2 



