260 Prof. G. Wiedemann on the Methods 



by the Committee of the British Association appointed for the 

 purpose in 1863. 



A wire coil is put into rotation about a vertical axis, in 

 consequence of which currents are induced in it, whose inten- 

 sity in unit time is, on the one hand, directly proportional to 

 the horizontal component of the earth's magnetism and to the 

 change of the projection of the plane of the coil on the ver- 

 tical plane at right angles to the direction of that component, 

 and, on the other hand, inversely proportional to the resistance 

 of the coil. 



But, then, the induced currents which traverse the coils act 

 upon the neighbouring coils, and induce in them extra currents 

 whose electromotive force is proportional to the change in the 

 directly induced current, in the unit time; hence the inductive 

 action in the spiral is diminished. The deflection of the needle 

 is determined by the total action of all the induced currents. 

 The following investigations are therefore necessary to deter- 

 mine the absolute resistance of the coil. 



1. The measurement of the space enclosed by the windings 

 of the coil, to which the induction is, cceteris paribus, propor- 

 tional, as well as its form, upon which the extra currents in- 

 duced in it depend, and the moment of rotation exerted on 

 the magnetic needle at its centre. 



These determinations offer very considerable difficulty. 



If thick wire is employed for Avindiug the coil, its diameter, 

 together with the insulating covering, must be exactly an 

 aliquot part of the interior width of the frame on which the 

 wire is to be wound, or else the outer layers of wire will be 

 squeezed more or less between the wires of the inner layers, 

 and so cause displacement. Moreover, as W. Siemens* has 

 shown, the wire is stretched in winding, the more the thinner 

 the wire is ; and this extension may be as much as 6 per cent. 

 Again, the insulating covering of the wire becomes pressed 

 together; this takes place less when the covering consists of solid 

 gutta-percha or similar substance than when the wire is covered 

 with silk or cotton. The thinner the wire is, the more im- 

 portant do these errors become, of displacement of the wire, of 

 extension on winding, and of the squeezing together of more 

 of the insulating covering in proportion to the diameter of the 

 wire. It is therefore not correct to calculate the space enclosed 

 by the coils from the length of the wire before winding or 

 after unwinding, and from the dimensions of the coil. 



On this account, as W. Siemens rightly remarks, the accu- 

 rate measurement of the length of the wire of about 1*1 milli- 



* PoggendorfTs Annalen, 1866, vol. cxxvii. p. 327. 



