ON ELECTJROLTSIS IN ITS PHYSICAL AND CHEMICAL BEAEINGS. 359 



conclusion that electrolytic and metallic conduction are essentially distinct 

 phenomena. 



Considerable discussion has taken place during previous years in the Committee 

 as to the theoretical determination of E.M.F. on the lijies originally indicated by 

 Sir Wm. Thomson from thermochemical data, and attention has been directed to the 

 necessity of taking into account reversible heat eftects. I may be allowed to re- 

 iterate the opinion that tlie evidence tending to throw doubt on Sir Wm. Thomson's 

 law connecting the E.M.F. of a cell with the attendant chemical changes cannot 

 as yet be regarded as in the least degree conclusive, and I venture to otter a brief 

 criticism on the experimental work which has been done in this direction — chiefly 

 by F. Braun (Wiedemann's 'Annalen,' 1882, 16, 561 ; 17, 593) and by Wright (' Pro- 

 ceedings of the Physical Society '), who have measured the E.M.F.'s of a large 

 number of different cells, and have compared the values found with those calcu- 

 lated from the chemical changes which they assume take place in them. The 

 calculated and observed E M.F.'s were found to agree in only a limited number of 

 cases ; in a considerable number the observed E.M.F. fell more or less below that 

 calculated ; in a few it was greater ; and Wright has specially instanced a number 

 of cells in which the current flows in the opposite direction to that which was to 

 be anticipated. 



The measurement of E.M.F. was made either by means of an electrometer ; or 

 "with a galvanometer, using currents of verij small intensity ; and there is no reason 

 to doubt the results. But both observers were, as I have said, content to assume 

 that the chemical changes "were of a particular kind, and made no attempt whatso- 

 ever to study them in detail from the chemist's point of view. 



Now it has been pointed out by Laurie (' Phil. Mag.' August 1886) that the 

 extraordinary results obtained by Wright with aluminium zinc cells, which give an 

 E.M.F. opposed to that calculated from the thermal data (of no less than 1"5 volt in 

 the case of the sulphates), are to be accounted for on the assumption that he was 

 not dealing with aluminium at all but with aluminium oxide supported on an 

 aluminium plate. It appears to me, then, that the whole of Braun's and Wright's 

 results are open to this criticism, as the method of measurement which they 

 adopted is one that would permit every surface impurity, every impurity in the 



' To reduce heat units per dyad gramme-molecule direct to volts, use the divisor 

 46,000. The column headed ' relative ' in the table would then run, 11, •35, 1'18 

 2-08, 1-80, 2-16, 1-00, -71, -97. 



