98 ELEMENTARY LESSONS ON [CHAP. 11. 



measuring magnetic repulsions. By means of the 

 torsion balance we may prove the law of inverse squares. 

 We may also, assuming this law proved, employ the 

 balance to measure the strengths of magnet poles by 

 measuring the forces they exert at known distances. 



To prove the law of inverse squares, Coulomb made 

 the following experiment : The instrument was first 

 adjusted so that a magnetic needle, hung in a copper 

 stirrup to the fine silver thread, lay in the magnetic 

 meridian without the wire being twisted. This was done 

 by first putting in the magnet and adjusting roughly, 

 then replacing it by a copper bar of equal weight, and 

 once more adjusting, thus diminishing the error by 

 repeated trials. The next step was to ascertain through 

 what number of degrees the torsion -head at the top 

 of the thread must be twisted in order to drag the 

 needle i out of the magnetic meridian. In the par- 

 ticular experiment cited it was found that 35 of torsion 

 corresponded to the i of deviation of the magnet; then 

 a magnet was introduced, that pole being downwards 

 which repelled the pole of the suspended needle. It was 

 found (in this particular experiment) to repel the pole of 

 the needle through 24. From the preliminary trial we 

 know that this directive force corresponds to 24 x 35 

 of the torsion -head, and to this we must add the 

 actual torsion on the wire, viz., the 24, making a total 

 of 864, which we will call the " torsion equivalent " of 

 the repelling force when the poles are thus 24 apart. 

 Finally, the torsion -head was turned round so as to 

 twist the suspended magnet round, and force it nearer 

 to the fixed pole, until the distance between the repelling 

 poles was reduced to half what it was at first. It was 

 found that the torsion -head had to be turned round 8 

 complete rotations to bring the poles to 12 apart. 

 These 8 rotations were an actual twist of 8 x 360, or 

 2880. But the bottom of the torsion thread was still 

 twisted 12 as compared with the top, the force pro- 



