262 Prof. G. Wiedemann on the Methods 



then we have, besides the sources of error of the original 

 apparatus, a number of other sources of error which need to 

 be specially examined; so that it is possible the accuracy of 

 the results might be seriously prejudiced. In any case the 

 difficulty of accurately determining the self-induction is the 

 most suspicious part of the method under consideration. 



2. The temperature of the coil must be determined with the 

 greatest accuracy, since the conductivity of the wire decreases 

 about 03 per cent, for a rise of temperature of 1° C. The 

 corresponding change of length, and consequently of surface 

 embraced, amounts only to tooVoooj ana ^ ma y therefore be 

 neglected. 



3. We have further to inquire whether there may not be 

 secondary currents induced in the supports of the apparatus, 

 if these are of metal, by the currents circulating in the spiral, 

 which may act upon the magnetic needle*. 



This point may be determined by interrupting the conti- 

 nuity of the metallic supports by means of insulating material. 

 According to experiments of this nature, made by Lord Ray- 

 leigh and Dr. Schuster |, this source of error was not important 

 in the experiments of the British Association, the error amount- 

 ing only to 0'16 per cent. It would be better, however, to 

 construct the supports of insulating material. 



4. The testing of the instrument to determine if the coils lie 

 symmetrically with reference to the axis of rotation offers no 

 particular difficulty, if we observe by means of a telescope cor- 

 responding points of the frame on both sides of the axis in 

 different positions of the coils, making with each other an 

 angle of 180°. 



5. In the same way it is easy to ascertain by known optical 

 methods whether the axis of rotation is really vertical, and 

 remains so ; deviation from the vertical position may exert a 

 considerable effect. 



If the angle of dip were about Z = 70°, an inclination of the 

 axis towards north or south of 0°'2 would cause an alteration 

 in the inductive action of the earth in the proportion of 

 cos 70°: cos (70°±0°-2)— that is, not less than 1 per cent. An 

 inclination of this amount in the case of the British- Association 

 coil would correspond to a displacement of 0*5 millim. in the 

 ends of the axis. An exact determination is therefore very 

 necessary. Small displacements of the axis towards east or 

 west have only an insignificant influence. 



6. The counting of the number of revolutions of the coil in 



* Compare F. Kohlrausch, in Pogg. Ann. 1874, Erganzungsband vi. 

 p. 9. 



t Loc. cit. 



