476 PROFESSOR W. E. AYRTON, MR. T. MATHER AJsD MR. P. E. SMITH: 



on starting or stopping the current was barely perceptible when the bronze rod was 

 inside the solenoid, and on removing the rod and opening the tertiary circuit, without 

 making any other change whatever, the balance was to all appearances perfect. We 

 were therefore certain that the permeability of the alloy differed very little, if at all, 

 from unity, so the casting of the stand was proceeded with. 



Similar tests were made on the completed stand when received at the Central 

 Technical College in 1900. The shape and size of the stand, however, made it 

 difficult to place within coils of manageable dimensions, so a modified method of 

 testing was used. The College possessed a standard of mutual induction (called S 

 in this section), of O'Ol henry, made in 1892, consisting of coils wound in grooves on a 

 wooden disc 9 inches diameter and 2f inches thick, so it was decided to test the 

 stand by observing whether the mutual induction of these coils was altered by 

 placing them on the circular top-plates of the slide rests which were to support the 

 coils. To do this, an induction balance formed of the mutual induction standard S 

 and another pair of coils was arranged as described above. The system was carefully 

 balanced when S was supported on one end of a pine* board, 1 inch by 11 inches by 

 12 feet long, the other end of which rested on one of the top-plates. On moving S 

 to the middle of the board the balance was not disturbed, but on placing it over the 

 stand a quick jerk of 85 divisions and rapid return to zero was noticed. This kick, 

 the effect of eddy currents in the metal of the support, was neutralised as far as 

 possible by a tertiary circuit. It could not, however, be entirely eliminated by the 

 tertiary coils available, a phenomenon attributed to want of equality in the time- 

 constants of the tertiary circuit and of the eddy-current circuits in the continuous 

 metal. The procedure adopted was to observe the swing produced by shunting 7^0 

 of the current from the primary of the balancing pair of coils, when the test pair were 

 supported above air, and on the top-plate of the stand respectively, the tertiary 

 circuit being open in the former case and closed in the latter. In each of the two 

 positions the swing produced was 33 divisions. The sensitiveness of the arrangement 

 was thus I in 33,000 per division, and under these conditions no difference would be 

 detected. Four sets of tests were made giving precisely equal swings. 



The experiments were repeated on the top-plate of the second slide rest of the 

 stand with the same result. The eddy-current effect was somewhat different in the 

 two cases, for in one the resistance in the tertiary circuit necessary to give minimum 

 kick was 134 ohms and in the other 124 ohms. 



To test whether the two ends of the pine board differed magnetically, it was 

 turned end for end, and the whole cycle of operations repeated. No difference could 

 be detected. In all cases great care was taken to twist the leads together in pairs, so 

 as to avoid mutual induction in parts of the circuit other than that under test. The 

 test coils S and the balancing coils were kept far apart and with their planes at right 



* Pine was used because previous work in connection with very sensitive moving-coil galvanometers 

 had shown this material to be non-magnetic. 



