364 Dr. C. Barus on the Relation between the Thermoelectric 



The minimum values ofT. E. H. (obtained by cooling the 

 red-hot bar as slowly as possible) of different kinds of steel* 

 and of soft iron are therefore approximately the same ; whereas 

 the maximum value of this constant (obtained by cooling the 

 highly heated bar as rapidly as possible) differ enormously, 

 this difference being a direct function of the quantity of carbon 

 contained, 



f Sir W. Thomson f has shown that in a thermoelement 

 consisting of magnetized and unmagnetized steel of the same 

 hardness and form, thermo-currents due only to magnetic 

 difference can be generated. The direction of these currents 

 was found to be in one case from unmagnetized to longitudi- 

 nally magnetized through warm, in another from transversely 

 magnetized to unmagnetized through warm, therefore also 

 from transversely to longitudinally magnetized through warm. 



For the purpose of informing myself of the magnitude of 

 the thermo-difference thus produced, the following experiment 

 was made : — A soft rod (I., Table II.) was tested for its 

 electromotive force when combined with copper (as described 

 above, p. 348), and the locus of the equation E T = cit—/3t 2 con- 

 structed from fifteen very carefully made observations. 



Hereupon a large magnetic battery weighing 40 lbs. was so 

 placed that each of the ends of the horseshoe touched a re- 

 ceiver. The distance between the poles of the magnet and 

 the corresponding ends of the steel rod was thus about 5 

 centims. A second series of fifteen observations was now 

 made. Upon comparing the locus of the latter with that of 

 the former, the two curves coincided so completely that no 

 influence could be discerned. Herefrom we conclude that 

 the thermoelectric effect due to a difference of magnetic state 

 may, in comparison with those which can be produced by a 

 difference in hardness, be completely neglected. 



g. A very curious analogy was found in comparing the re- 

 sults at which FrommeJ arrived, in studying the specif c gra- 

 vity of differently tempered steel, with the T. E. H. of similar 

 rods as found in my experiments. Dr. Fromme, if I infer 

 correctly, limited his experiments to rods whose diameter was 

 greater than 2 millims. and less than 7 millims. His results 

 are contained in the following Table, the volume of the soft 

 bar being put = unity. 



* Steel is here used as distinguished from iron only by containing a 

 greater proportion of caribou. No attention has been paid in this paper to 

 the effects of P, S, Si, Mn, &c, so often present in both. 



t Phil. Trans, iii. pp. 722-727, 1856. 



I Fromme, Gotting. Nachr. 165, 1876. 



