MAGNETIC PERMEABILITY OF IRON, STEEL AND XICKEL 41 



specially constructed for the purpose. It was on the principle of Thom- 

 son's reflecting instrument, but was modified to suit the case by increas- 

 ing the size of the mirror to of an inch, by adding an astatic needle 

 just above the coil without adding another coil, by loading the needle 

 to make it vibrate slowly, and, lastly, by looking at the reflected image 

 of the scale through a telescope instead of observing the reflection of a 

 lamp on the scale. The galvanometer rested on a firm bracket attached 

 to the wall of the laboratory near its foundation. In most of the ex- 

 periments the needle made about five single vibrations per minute. 

 The astatic needle was added to prevent any external magnetic force 

 from deflecting the needle; and directive force was given by the magnet 

 above. Each division of the scale was 075 inch long; and the extrem- 

 ities of the scale were reached by a deflection of 7 in the needle from 0. 

 The scale was bent to a radius of 4 feet, and was 3 feet from the instru- 

 ment. At first a correction was made for the resistance of the air, &c. ; 

 but it was afterwards found by experiment that the correction was very 

 exactly proportional to the deflection, and hence could be dispensed 

 with. This instrument gave almost perfect satisfaction; and its accu- 

 racy will be shown presently. 



The tangent-galvanometer was also a very fine instrument, and was 

 constructed expressly for this series of experiments. The needle was 

 1*1 inch long, of hardened steel; and its deflections were read on a 

 circle graduated to half degrees, and 5 inches in diameter. The aver- 

 age diameter of the ring was 16^ inches nearly, and was wound with 

 several coils; so that the sensibility could be increased or diminished 

 at pleasure, and so give the instrument a very wide range. The value 

 of each coil in producing deflection was experimentally determined to 

 within at least ^ of 1 per cent by a method which I shall soon publish. 

 The numbers to multiply the tangent of the deflection by, in order to 

 reduce the current to absolute measure, were as follows: 



Number of coils. Multiplier. 



1 -05377 



3 -01800 



9 " . -006007 



27 -002018 



48 " . -001143 



By this instrument I had the means of measuring currents which 

 varied in strength several hundred times with the same accuracy for 

 a large as for a small current. For greater accuracy a correction was 



