2 Dr. F. Auerbach on the Passage of 



{'2) The conductivity diminishes, or the resistance increases, 

 as the temperature rises. If we put 



\, = \ (l-«*+[a 2 + /3]* 2 ) 

 and 



?Q = ?r (l +0^-/3/), 



then is, at all events very nearly, 



a. Y — oi and ^ = 0. 

 Therefore 



/S 1 = in case = 0. 



This latter relation was found by E. Becquerel, Arndtsen 

 (approximately), and Mousson. At the same time 



Becquerel gives a = 0*004726 

 Arndtsen . . 0'00413 

 Mousson . . 0-004207 



On the contrary, Matthiessen finds /3j_ different from 0; 

 namely, from 



x= X o (l-0-0051182* + 0-000012915£ 2 ) 



folloAVS 



lv = lVo (l + 0-0051182 £-0-000013281 * 2 ). 



But here also we can, with very close approximation, regard 

 the formula 



iv = ic Q (l + ctt) 



as satisfied. This assumption, namely, leads to the equation 



\ = X (1— a* + «V — ...) 

 or 



x=x A l -" t+ T+«[ty> 



in which [t] signifies a mean value of t, which can be intro- 

 duced, in that term of the correction, in place of the true one. 

 It is true that the statements of the observer furnish no cer- 

 tain support for this ; but if it be put, in a round number, equal 

 to 200° C, we get 



/ «¥\ 



X = X ( 1 — ctt+ -a~)j 



as in fact Matthiessen found (nearly). 



(3) The quantity of heat generated by the current in an 

 iron wire is approximately determined by Joule's law. I have 

 not been able to discover any numbers referring to this more 

 accurate than the old ones of Lenz, which, on account of the 

 temperature rising with the intensity of the current, are not 



