Dr. (J. (Jorv. Relation* of Heat to Voltaic and 



Table III (continued). 



Metal. Liquid. 



Mg KCy 



Pt Tartaric acid 



Ag K alum 



Pd HC10 3 



Mg KHO 



Pt 



Mg 



Total 1 -0057 



Average '0670 



Remarks. The current from the strongest pair in this series had 

 about one hundred times greater potential than that from a single pair 

 of bismuth and antimony with an equal difference of temperature. 



Not only were the thermo-electro-positive combinations of metals 

 and electrolytes about double the number of the negative ones (see 

 Table I) : but the range of potential of the strongest of the former 

 was about 4'12 times that of the latter, as shown by this table. The 

 facts that there is a much larger number of thermo-electro-positive 

 metals than of thermo-negative ones ; also of thermo-electro-positive 

 combinations of metals with liquids than of negative ones, and that 

 the positive elements of each of these classes are usually the strongest, 

 indicate that electro-positive action in metals generally is more fre- 

 quently increased than decreased by rise of temperature. 



In the present table, the acids in contact with the chemico-electro- 

 negative metals platinum, gold, and palladium, are thermo-electro- 

 positive ; and in contact with the positive ones, aluminium and 

 nickel, they appear negative. In this table also, as in Table I, 

 aluminium, a strongly chemico-electro-positive metal, is conspicu- 

 ously the most thermo-electro-positive one. The highly chemico- 

 electro-negative metals also, silver, palladium, gold, and platinum, are 

 present only in the thermo-negative division. In apparent opposition 

 to this, bat still largely in accordance with Table I, it is remarkable 

 that magnesium, a metal highly chemico-electro-positive, and also 

 considerably thermo-positive in the series of metals only, is the most 

 thermo-negative in this series (see also pp. 255, 261, 270, and 278). 



In order to ascertain the influence of strength of liquid upon the 

 sequence of the various thermo-electric series, another set of experi- 

 ments, similar to those of Table I, was made, with all the solutions of 

 Hve times the strength, except those of sodic diphosphate, which was 

 four times, and those of potassic and ammonic alum, and potassic sul- 

 jihate, each of which was only three times. The results are shown in 





