the Galvanic Current through Iron. 221 



by an equal quantity of beat. Tbe rise of temperature tbus 

 amounts to o, 8 R. or i° C. Six seconds were required for tbe 

 determination of tbe resistance. As, then, a rise of tempera- 

 ture of 1° C. augments the resistance 1 by 0'0005, thermal 

 influence appears in fact not to be excluded. 



These reflections I first made after having, as I believed, 

 already proved, by numerous series of experiments without 

 taking this circumstance into account, the dependence of the 

 resistance upon the intensity of the current. Of the results 

 of some of those series I will quote the mean values, — first, 

 because at least so much can be inferred from them, that even 

 after deducting the thermal alterations of resistance others 

 still remain, and, secondly, because the comparison with later 

 experiments, which were free from thermal influences, fur- 

 nishes indeed a confirmation of the calculation we have just 

 made. 



1. Unannealed iron wire. £ = 2120, cZ=0 # 28. In the mean 

 of symmetrically distributed experiments in sets of 5 each: — 



(ID) ti* =7-7570, (2D) w> 3 =7-7600, (3D) «- 3 = 7-7636; 



2^-^ = 00030, w 3 r-w 2 =0-0036. 



Therefore w n +i~w n is nearly constant, and 



w n+1 — w 



-=8 = +0*00043. 



This number, it is true, is below the extreme value found 

 for the effect of the heating ; but the probable value corre- 

 sponding in this case to the extreme value of that effect does 

 not at any rate amount to more than the half of the former ; 

 it therefore amounts to no more than 0*00025. It is there- 

 fore established, to a high degree of probability, that the resist- 

 ance increases with the intensity of the current. 



2. Repetition of experiment 1. 



2^ = 7*7482, 1^=7-7526, i0 3 =7-7574; 

 therefore mean of S = 0*00059. 



3. As experiment 2 had given a higher value of 8 than ex- 

 periment 1, so a third series of experiments gave for it a still 

 higher value ; for the result was: — 



Wl = 7-7469, w 2 = 7-7528, w- 3 = 7*7587 ; 



8 = 0*00076. 



This gradual increase of 8 furnishes a confirmation of what 

 was assumed at p. 219, that the high values of 8 in the wire 

 there under investigation might be a consequence of its having 

 been several times used galvanically. 



