270 Prof. G. Wiedemann on the Methods 



We have therefore for these determinations to ascertain 

 the dimensions (1) of the wire ring, (2) of the inducing coil, 

 (3) of the inducsd coil, (4) of the multiplier, (5) of the dis- 

 tance of the separate layers of the inducing and induced coils 

 from each other, (6) the determination of the position of the 

 magnetic poles, to which are to be added the manipulations 

 required for the other methods, the adjustment of the needle 

 in the centre, and of the coils of the multiplier and of the 

 wire ring between them with reference to the meridian &c. 

 Since the intensities of the primary and secondary currents 

 are measured by the deflection of the same magnet, the hori- 

 zontal component of the earth's magnetism does not enter into 

 the calculation. There are in this method many more sources 

 of error to be taken into account than in the other methods, 

 by which its accuracy may be seriously impaired. 



Notwithstanding, Weber finds, by determining the resist- 

 ance of the inductive circuit by this method, and comparing 

 it with the Siemens standard which he employed, that at two 

 different distances of the inducing and induced coils, and with 

 two different intensities of the inducing current, a mean result 

 of 0-9554 (0-9589 to 0-9516)*. 



This method has also been employed by Rowlandt,onlv that 

 he employed a tangent-galvanometer to determine the inten- 

 sity of the inducing current, but a special galvanometer to 

 determine the induced current ; it was necessary also to deter- 

 mine the ratio of the intensities of horizontal force at the posi- 

 tion of the two pieces of apparatus. His three induction- 

 coils had a diameter of about 27*4 centim., and were wound 

 on frames having a thick brass rim on each side. They could 

 be combined in pairs, so that the one which served as inductor, 

 through which a constant current flowed which could be in- 

 terrupted, might be placed at four different distances from 

 the others (6*5 up to 11*47 millim.). Very thin-covered 

 copper wire (no. 22) was taken for the coils; so that the 

 sources of error mentioned on p. 260, caused by displacement 

 of the covering, the mutual compression of the coils, and ex- 

 pansion of the wire, become very prominent. Since the in- 



* An indirect method employed by Fr. Weber to determine the resist- 

 ance of a zigzag platinum wire placed in a calorimeter, from the heat pro- 

 duced by a current -whose strength in absolute measure was determined 

 by means of a tangent-galvanometer of known dimensions, can hardly be 

 placed by the side of the more direct methods, on account of the difficulty 

 of all calorimetric measurements. Nevertheless Fr. Weber finds by this 

 method, by comparison with his Siemens standard, a resistance for the latter 

 of 0-9560, which again differs only - l per cent, from the other determi- 

 nations. 



t Rowland, * Silliman's Journal,' 1878 (3), pp. 325, 430. 



