Thermo-electric Action of Metals in Electrolytes. 287 



positive or less negative to that in which it was less positive or more 

 negative, was 384 to 199 (= 65'87 per cent.), or as 1'93 to 1 ; also by 

 similarly comparing Table XX, that the proportion in which the 

 metal most corroded at 160 F. was most positive or less negative to 

 that in which it was the reverse, was 338 to 161 (= 69'12 per cent.), 

 or as 2 - 24 to 1. * 



And with regard to the relation of corrosion to voltaic state, it has 

 been assumed " that in all these cases the positive electricity sets out 

 from the more oxidable metal, and traverses the liquid towards the 

 less oxidable one;" but comparisons of Tables X, XIX, and XX 

 show that this is not entirely correct ; with twelve different liquids at 

 60 F. the most chemico-electro-positive metal was also the most 

 corroded one in 337 cases, and the least corroded in 62 =a proportion 

 of 5'43 to 1, or 84'44 per cent. ; and with seven liquids at 160 F. the 

 numbers were respectively 248 and 59= a proportion of 4'20 to 1, or 

 80" 77 per cent. Neither volta-positive nor negative state, therefore, 

 is always proportionate to degree of ordinary corrosion, either in cold 

 or hot liquids. These figures show also that by a rise of temperature 

 of 100 F. the proportion of cases in which the negative metal was 

 the most corroded, increased from 15*56 to 19'23 per cent, of the 

 entire number of cases. Comparison of the above proportions with 

 those of the relation of corrosion to the thermo-electric states of 

 metals in liquids, viz., 1'93 to 1 and 2'24 to 1 (see above), shows 

 that corrosion usually influenced in a much greater degree the 

 chemico- than the thermo-electric properties of such combinations. 



Not only was the relative number of cases in which the volta-nega- 

 tive metal was the most corroded increased by rise of temperature, 

 but also the average relative loss by corrosion of the negative one to 

 that of the positive was more than doubled. 



- The following appear to be cases in which contact influenced the 

 thermo-electric currents of metals in liquids. With each instance, in 

 the first part of the experiment the metal was suddenly immersed in 

 the hot liquid; and, in the second part, the same metal and liquid 

 were gradually heated. In a solution of potassic bromide, copper 

 alone was thermo-positive, but when gradually heated in contact with 

 tin it was thermo- negative ; in one of potassic nitrate platinum alone 

 was thermo-positive, but in contact with gold it was rendered slightly 

 more negative during a slight elevation only of temperature; in 

 potassic hydrate magnesium alone was thermo-negative, but in con- 

 tact with lead it was rendered more positive by heat up to 170 F. ; 

 in sodic sulphate nickel alone was thermo-positive, but in contact 

 with tin it was more negative up to 110 F. ; in dilute nitric acid 

 lead alone was thermo-positive, but in contact with iron it was more 

 negative up to 180 F. ; in the same liquid tin alone was thermo- 

 negative, but in contact with nickel it was more positive up to 160 F. ; 



