10 Dr. F. Auorbach on the Passage of 



therefore Por its consequence a momentary but no permanent 

 change of sign of 8. That the repeated opening and closing 

 of the magnetizing current exercised no influence depended 

 probably on stationary relations having already entered, in con- 

 sequence of the wire having been long in use. In order to 

 test this supposition I gave to a yet unused iron wire of like 

 constitution the same dimensions, and obtained: — 



12. [> ] = 6018 6020 6017 6015 6013 

 [«,]== 6011 6011 6008 6006 



Means:— M = 6016-6; [>] =6009-0; [w-Wq] = -7'6; 



8 = -0*0032. 



Here, therefore, 8 is negative. In fact, after I had closed 

 and opened the magnetizing current about 100 times, the ab- 

 solute value of 8 diminished, and on the next day I obtained, 

 as in 10 and 11, a small positive value of 8. But I have only 

 observed such change of sign of 8 when, for n = 92, the ratio 

 H : M = l : 2 or did not much differ from this. Only when 

 very intense currents had passed through the wire for a longer 

 time, or when I had repeatedly reversed the direction of the 

 magnetization, did the results become in other cases also and 

 completely irregular. 



Shortly before the construction of the adiathermanous ap- 

 paratus, I made some more experiments, in which, it is true, 

 I verified an influence of the heating ; but I believed I should 

 be able to eliminate it by comparison with experiments on 

 copper wires. The specific heats, namely, of iron and copper 

 are to one another about as 7 : 6 : and, according to Benoit, 

 their temperature-coefficients a are approximately in the same 

 ratio for the galvanic resistance. If, then, we employ the 

 same value of M and the same magnetizing-apparatus, and 

 give to the surfaces of the iron and the copper wire equal mag- 

 nitude, by which also their A^olume becomes the same, we get 

 in both nearly equal augmentations of resistance by radiation 

 from the magnetizing-apparatus. Hence I proceeded in the 

 way indicated, and obtained (the indices / and c refer to the 

 soft-iron and the copper wire respectively) : — 



13. Resistance-box. H=1D; M = 2B; n=73. 



8 C = 0-0023 8 f = 0-0042 

 o>— S C =S= +0*0019. 



14. The same repeated. 



8 C = 0-0026 6>= 0-0039 8= +0*0013. 



15. H = 1D; M = 3B: ra=73, thicker wires. 



a c =0-0017 5>= 0-0044 8= +0*0027. 



