arid the Theory of the Induction-balance. 145 



a tone mixed up with it which is shriller than any produced 

 by the sonometer. (I do not know whether this has ever been 

 noticed experimentally. One often notices a change of pitch 

 in the telephone-rustle ; but it appears to depend in some 

 cases on the direction of the current, i. e. on whether it 

 strengthens or weakens the magnet.) 



If the sonometer is to be used, I would suggest that the 

 second primary coil be done away with, and that the secondary 

 coil be of small diameter. It may have as much wire on it as 

 is wanted for long range ; but the further it is away from 

 the primary, compared with its own diameter, the better. The 

 diameter of the primary does not much matter. The in- 

 duction would now simply vary as the inverse cube of the dis- 

 tance from the mean circumference of the primary ; hence, if 

 the sonometer were graduated so as to give equal distances 

 from this line, its readings would only have to be turned 

 upside down and cubed to give comparative results. If both 

 primary and secondary coils are small, the sonometer-arm 

 may be graduated like an ordinary millimetre-scale. 



If an absolute zero were wanted at any time, it might be 

 obtained by an arrangement for rotating one of the coils till 

 it was at right angles to the other. 



22. Prof. Hughes's second method of measuring, that of 

 the graduated zinc wedge, is a much better one, because it is 

 a null method and gives true readings, though they are not 

 easily interpretable. 



A third method has, I believe, been tried, viz. a copper 

 damper rotating above one pair of coils ; and this seems also 

 pretty good ; but I think it might possibly be better to modify 

 it by having a uniform disk (or ring) of high-conductivity 

 copper capable of being moved along the axis of one pair 

 of coils with its plane always parallel to theirs. The distance 

 of the disk from the middle point of the coils should be read, 

 and the distance apart of the coils should be constant. Given 

 all particulars, I have no doubt that a mathematician could 

 readily interpret results so obtained. 



Another modification would be to use a coil of wire form- 

 ing a circuit closed within itself instead of the copper disk. 

 It would have just about the same effect as a solid disk of the 

 same size ; but the data connected with it might be more 

 definite. 



The Effect of Magnetic Bodies. 



23. If a thin piece of iron wire be held in the balance, it 

 powerfully disturbs it by a concentration of the lines of force 

 in the iron, so that M (the mutual induction between the 



Phil. Mag. S. 5. Vol. 9. No. 54. Feb. 1880. L 



