the Galvanic Current through Iron. 



225 



y\ both will approach to the horizontal rectilineal form for x 

 increasing ; but this right line will possess a larger ordinate 

 for hard wires, and commence at a larger abscissa. Compare 

 fig. 8. 



9. Repetition of 7. w= 102-4. 



n. 



a+b. 



s. 



s'. 



w. 



to'. 



Mean. 



Difference. 



1 



3-4 



20 r 



2 9 r 



102-341 



102-315 



102-328 



16 

 12 

 14 

 16 

 18 



2 



7-8 



3-2 



5-5 



359 



329 



344 



3 



120 



3-6 



7-0 



370 



342 



356 



4 



15-9 



'25 



6-9 



384 



357 



370 



5 



18-9 



0-2 



53 



399 



373 



386 



6 



22-0 



3-3 1 



1-2 



415 



394 



404 



In all qualitative relations this series agrees with 7. The 

 differences decrease very slowly ; from n = 4 they increase, in 

 consequence of thermal influences ; only the absolute values 

 of the differences are somewhat greater; namely, there becomes 



S x 2 = 0*00016. 



In order to fix for steel also the curves drawn for hard and 

 soft iron, I made experiments with wires of elastic steel. They 

 did not, however, in general yield results of corresponding 

 trustworthiness. Only thus much could be concluded from 

 them, that the ordinate of the straight horizontal line in which 

 the resistance-curve with increasing intensity of the current 

 ends is much greater than with hard iron, and commences at 

 a still greater length of abscissa. I will communicate at least 

 one such series of experiments. 



10. Steel wire F 8 . Z = 3500, d = 0'A3. Approximately 

 10 = 3-53 (for s) and =3'54 (for /). 



01. 



a-\-b. 



s. 



s[. 



w. 



w'. 



Mean. 



Difference. 



1 



8 



3-7/ 



5-8 



3-5346 



3-5328 



3-5337 



10 



11 



? 4 



10 



8 



2 



14 



8-3 



93 



59 



35 



47 



3 



19 



11-6 



8-6 



61 



55 



58 



4 



24 



13-5 



7-6 



?56 



68 



?62 



5 



28 



19-9 



7-6 



71 



73 



72 



6 



31 



25-8 



7-0 



83 



77 



80 



Therefore 



8 12 = 0-00033. 



But the total alteration is not yet concluded, even at n = ti. 



The following last two Tables refer to control-experiments 

 with copper wires. In both, for b\ 2 a very small negative 

 value came out ; at greater values of n the influence of the 

 heating then asserted itself. 



