the Galvanic Current through Iron, 13 



24. Repetition of 22. As mentioned, it gave an augmen- 

 tation, though hardly sensible, of w . For a control I pro- 

 duced, out of the same coil of wire from which / 9 had been 

 taken, a new wire of the same dimensions, nearly, and found 

 for H = 2D, M=1D (therefore as in 22):— 



w = 7-3608 w= 7-3559 w = 7-3602 



8= -00006, 



therefore a negative value of 8, as in 22. 



25. In experiment-series 6 and 21 the conclusion was reached 

 that the resistance of hard iron wires was not changed by 

 magnetizing them. I sought to realize the same case with a 

 tempered thick steel wire F a (1= 1450, d— 1*31). I succeeded, 

 when H was =2D and n = 92, with almost equal approxima- 

 tion when M was =2D and M = 3D. This point can be fixed 

 still more exactly with annealed iron wires. On the contrary, 

 in hard steel wires there was always an alteration of the resist- 

 ance; and, indeed, with H constant, the more insignificant 

 the magnetizing force the smaller in general was the altera- 

 tion. Of the experiments with slight magnetizings I will only 

 mention that they mostly yielded lower values of 8 than those 

 with iron wires ; only with the lowest values of the ratio Mw : H 

 does the opposite take place. On the other hand, I will give 

 some more series of experiments with magnetizing forces to 

 which, in iron and tempered steel, positive values of 8 would 

 have corresponded. 



26. H = 1D 5 M = 3B; n=110. Wire F 3 of English steel 

 (music-steel). £=1280; d = 0'91. Mean values (the first 

 experiments being excluded) : — 



■*+{?=S2079}-" »=-°-° 0012 

 (w ft = 0-050 nearly). 

 8= -000034. 



If now the magnetizing current was reversed, there followed, 

 as at the commencement of the first series, first an increase of 

 8 by magnetizing, which, however, after the current had 

 passed twice had already given place to a diminution. Gene- 

 rally, this wire proved very favourable, not only for these 

 measurements, but also for those of which I have yet to speak 

 in the course of this investigation. But the absolute values 

 of 8 are here much smaller than in other wires; already on the 

 25th May, with a series (M) of 5B, I had obtained scarcely 

 higher values of 8. 



In order to determine whether the cause of this was the 



