356 Dr. 0. Barns on the Relation between the Thermoelectric 



K' K" 



-K- and -K- were determined in the same way previous to 



the experiments, and their values checked from time to time. 

 In comparing the resistances I. . . . IX. and A . . . D, the tol- 

 iowing plan was observed : — 



W:£... 



I 



IV 



IV 







VIII 



VIII 



I 



1 And in the same way 

 with A . . . D. 



R: d ... 



II 



II 



V 





VII 



IX 



IX 



Each of the data is therefore derived as a mean of four deter- 

 minations. Possible heterogeneity of the wire a + b thus 

 becomes less effective. 



As an example, I will add results obtained for the rod C 

 (resist. =0'0051 Siem. U.):— 







Compared 



with 



A. 



Compared 



with 



B. 



Compared 



with 



D. 



1st position 



W :S = 



0-1648 



0-1654 



0-1649 



2nd position 



W:d = 



01630 



0-1632 



0-1639 





Mean 



0-1639 



0-1643 



01644 



For the rod I (resist. = 0*0542 S. U.) on different occasions, 

 1-735 and 1*741 were found for W : 8. 



As will be seen, the principle stress was put on the relative 

 values of the resistances. To facilitate orientation, however, 

 the results were approximately reduced to Siem. units by de- 

 termining the standard 8 in that denomination. 



Assuming the coefficient for temperature for steel to be the 

 same as that for copper, the results obtained will be good for 

 0° C. directly, 8 having been previously reduced. Although 

 this is only approximately true, the influence of this difficulty 

 on the relative values of the resistances will be but slight, as 

 the temperature of the room remained nearly constant. 



Table IV. 











AS 



No. 



Resistance. 



S. 



AS . 



T. E. H. ' 



I 



0-05417 



0-1361 



Nil. 







II 



0-04305 



01400 



00039 



1260 



*IV.* ... 



0-11130 



02337 



00976 



1990 



V 



0-11060 



0-2483 



01122 



2020 



VI 



0-07846 



0-2592 



0-1231 



2030 



VII 



08740 



0-2648 



0-1287 



2010 



VIII 



0-15330 



0-2779 



0-1418 



2050 



*IX 



0-09417 



0-2810 



01449 



2070 



1 The resistances marked with an asterisk were determined by using 

 the magneto-inductor as current-generator. In the others a Smee element 

 was employed. 



