360 REPORT— 1888. 



agents — not excluding dissolved oxygen — to exercise its proper effect. Julius 

 Thoinsen has expressly recognised this source of error, and has pointed out the 

 absolute necessity of measuring the E.M.F. of the element in full activity if it be 

 desired to determine the full effect produced by the chemical interchanges which 

 are involved in its action (' Thermoch. Untersuchungen,' III., p. 487). On this 

 account his comparisons of observed and calculated E.M.F.'s may well be set against 

 those of Braun and Wright ; they are as given on the preceding page. 



When it is borne in mind that in the case of nitric acid the products of change 

 gradually accumulate and produce an effect, the agreement between theory and 

 practice is remarkable. 



Braun argues that ' chemical energy ' is of the form of heat energy, and that iu 

 its conversion into electrical energy in the cell there must therefore necessarily be 

 some loss — some degradation ; he terms that fraction of the total energy which 

 takes the form of current the electromotive efficiency (Helmholtz's Freie Energie), 

 and seeks to connect this with the dissociation temperature of the compounds 

 formed at the electrodes. But if, to use Faraday's words, ' the forces termed 

 chemical affinity and electricity are one and the same,' there is, I imagine, no a 

 priori objection to be urged against Sir Wm. Thomson's law. 



Braun has based a determination of the 'electromotive efficiency ' of a number 

 of chemical changes on the assumption that the E.M.F. developed, for example, 

 between silver, silver bromide, bromine and platinum, is to be regarded as developed 

 simply in the formation from silver and bromine of silver bromide : the value found 

 was 84 ; that deduced from the heat of formation is 91 ; so the electromotive effi- 

 ciency is — = 'OS. The experiments were made in a very careful manner, and excep- 

 tional precautions were taken in purifying and drying the materials, the bromine 

 used being purified by Staas's method. But the fact that a current wan produced 

 between silver, bromine and platinum, and between silver coated with silver bromide, 

 bromine and platinum is, to my mind, sufficient to throw doubt on the results. It 

 is impossible to prepare pure bromine ; and it is a fair assumption to make that, if 

 the materials had been pure, no current would have been ob.'ierved : the determina- 

 tion of the E.M.F. is therefore of uncertain value ; the number deduced cannot have 

 been the outcome simply of the formation of silver bromide, as secondary change 

 must have taken place to some extent. 



I therefore venture to think that, up to the present, no experimental disproof of 

 Sir Wm. Thomson's generalisation has been given: on the other hand, the fluctua- 

 tions in E.M.F. which accompany slight changes in composition — such as have been 

 so clearly brought out by Lord Rayleigh's investigation of the Clark cell and by 

 Wright's and Fleming's investigations of the Daniell cell — and also the influence 

 which every change in the conditions exercises on the course of chemical change, 

 would appear to favour the assumption that the electromotive efficiency of chemical 

 change does not differ from the theoretical to the extraordinary extent that Braun 

 suggests. 



Before the temperature of dissociation can be taken into account as suggested 

 by Willard Gibbs, we require to know to what extent this depends on the nature 

 of the surface in contact with the dissociating substance. 



It is also desirable that a clearer statement of the nature of the reversible heatt 

 effects should be given. I can picture to myself that on passing a current across a 

 metallic junction the molecules become somewhat deflected from their original posi- 

 tions, or rotated, and that heat is developed on their subsequent return towards the 

 normal position ; but if this be the nature of the Peltier effect, and the con- 

 verse action occur when thermo-electric currents are developed, it does not appear 

 probable that the reversible heat effects at metal-liquid and liquid-liquid junctions 

 would be of the magnitude to account for the extraordinary differences said to 

 obtain in many cases between observed and calculated E.M.F. 



