226 



Dr. P. A.uerbach on the Passage of 



11. Wire of pure copper. /=? 18000, <£=0'41, w (approxi- 

 mately) = 4*32 and 4*31 respectively. 



n. 



a+£. 



5. 



s\ 



to. 



w'. 



Mean. 



Difference. 



1 



6-8 



20 r 



0-4 J 



4-3171 



4-3106 



4-3138 



-2 



+2 

 

 1 

 1 



2 



11-2 



3-6 



06 



68 



05 



36 



3 



15-0 



4-7 



1-0 



69 



07 



38 



4 



197 



6-1 



13 



69 



07 



38 



5 



24-0 



7-0 



1-6 



71 



07 



39 



6 



27-8 



7*7 



2-1 



72 



08 



40 



8 12 = -0*00005. 

 12. Platinized copper wire. 1= 15000, d=0-12, *tf = 30'15. 



to. 



a+b. 



s. 



s'. 



w. 



w'. 



Mean. 



Difference. 



1 



1-3 



0-0 



0-21 



30-1500 



30-1515 



30-1507 



-16 

 + 3 

 +12 



26 

 40 



2 



2-2 



0-2 r 



0-2 r 



1491 



1491 



1491 



3 



2-7 



0-2 



o-i 



1493 



1496 



1494 



.4 



3-0 



0-1 Z 



0-3 1 



1503 



1510 



1506 



5 



3-2 



0-9 



1-1 



1529 



1534 



1532 



6 



3-3 



2-3 



2-5 



1569 



1576 



1572 



8 12 = -0 00005. 



I have shown that, when the intensity of the current rises 

 from ID onwards, the resistance also increases ; and with soft 

 iron wires we could follow this increase up to its limit. But 

 now, how does the resistance change when, starting from ID, 

 we let the intensity fall ? That then w also declines is indu- 

 bitable; but two circumstances, further, favour the conclusion 

 that it sinks rapidly, at least in soft iron wires. For, first, we 

 have seen that the quantity w n+l —w„ diminishes when n in- 

 creases, and therefore increases with a falling n ; and there is 

 no ground for assuming that the curve changes its law at the 

 arbitrary value # = 11). And, secondly, the magnetizing- 

 experiments give us direct information respecting the total 

 change of resistance with the circular magnetizing, conse- 

 quently with the current-intensity : the negative values of 8, 

 however, were disproportionately greater than the positive 

 values which were here given for 5 14 ; the rest must therefore 

 arrive at the quantity S 01 . I have tried various methods for 

 determining this quantity — that is, for comparing the resist- 

 ance of an iron wire when the current is infinitely little with 

 the resistance when the current has a finite intensity. They 

 have, for the present, all failed, partly at the limit of sensitive- 

 ness of the galvanometer, partly at the comparison, which 

 could not be carried out, even when the absolute determina- 



