Soft Iron, especially with weaker decomposing -powers. 55 



by w b . This resistance was determined by Thomson's method *, 

 with only a slight modification, which was required by the cir- 

 cumstance that I had not to measure off two resistances accord- 

 ing to a given ratio, but to compare each two already definitively 

 measured and prepared resistances. In this manner, for the four 

 bridges used, the resistances w b at 20 o, 4 C. were found to be 

 respectively equal to 



1-1626 xlO 8 — , 22589xl0 8 — \ 



sec. ' sec. » 



1-1683 x „ „ 2-2811 x „ „ ; 



so that by using the two small wires in juxtaposition a resistance 

 is obtained which amounts to less than 0*002 of the least value 

 of SMB. 



In all other respects the method of measuring was identical 

 with the one first described : now the first deflections of the 

 magnet were observed ; then the multiplication method was em- 

 ployed. The disturbing influence of the extra currents (which 

 now proceeded only from the wire P), as well as of the interrup- 

 tion of the current, tends always to diminish the deflection of the 

 magnet, which is occasioned by the primary current and has to 

 be calculated from multiplied readings. This influence it is easy 

 to eliminate ; the mean value, however, of the induction-deflec- 

 tions is not affected by it, if we always carry out the two shift- 

 ings of the commutator Cj. 



I give also an example of this second method : — 



September 3, 1871. One Daniell; ?i = 800; secondary wire 



= division of 100 turns. Consequently — - = 688*30 (w s is 



W}y 



the resistance of the branch B M S B, fig. 4). 



In B the two smaller bridges were inserted in juxtaposition; 



temperature = 21 0, 2; consequently 1^ = 5 -8687 x 10 7 j^ • 



Further 



T = 20'640, \=0'16571, H = 2-0018, D = 2281 sc. 



The multiplications (10 readings each) gave as positions of 

 equilibrium of the magnet, 



409-80, 584-67, 58478, 409-88,— 



and as the values of the induction- deflection, 



124-23, 122-83, 122-61, 123-90. 



Correction for extra currents =1*81. 

 From this we find 



a = 89-33 sc, A =123-40 sc, 



* Phil. Mag. S. 4. vol* xxiv. p. 1-10; Wiedemann, GalvanismuS) vol. ii. 

 p. 1046* 



