3 4 SCIENCE PROGRESS 



wire are wound upon flat rectangular frames. They are 

 arranged upon a stand one above the other, so that in two of 

 them the wire is perpendicular to that in the remaining two ; 

 and each pair composes two opposite branches of the Wheat- 

 stone bridge arrangement. The battery and galvanometer are 

 put in in the usual places ; and adjustment for no current 

 through the galvanometer is secured by means of a slider which 

 connects, through it, the junctions of adjacent arms. 



The intended method of procedure, which it has been 

 necessary, however, to modify in actual practice, is as follows : 

 Let the stand be placed, at the time when the ether drift is 

 horizontal, so that two coils are parallel and two perpendicular 

 to it, and adjust the slider so that there is zero current in the 

 galvanometer. Now rotate the stand through a right angle, so 

 that the wire which was parallel is perpendicular. If any 

 change of resistance is consequent upon this rotation, the 

 balance will be destroyed, and a current will flow. It can be 

 easily shown that a change of resistance equal to that previously 

 calculated would, neglecting the resistance of the battery, pro- 



duce a current in the galvanometer of magnitude R \_ , where 



E is the electromotive force of the battery, g the resistance of 

 the galvanometer, and R the resistance of each coil. In the 

 present apparatus R is n ohms and g 10 ohms; the battery 

 used is a single storage cell, so that E is about ri volts. The 

 current to be expected, therefore, at a time when /3 = io -4 , is 

 about io ~ 9 ampere. The sensitiveness of the galvanometer is 

 such as to give almost exactly i centimetre scale deflection with 

 this current ; and, provided that disturbing causes are removed, 

 it becomes quite easy to say whether such a current exists or 

 not. But it is an extremely difficult thing to eliminate these 

 disturbances sufficiently to make a determination possible. 

 Their nature and the means of getting rid of them will be de- 

 scribed in detail in the proper place; here, it will suffice to 

 mention only one or two of them. One of the greatest difficulties 

 is the changes of resistance which occur owing to unequal 

 heating of the various wires. This is, no doubt, due in part to 

 draughts of air, and in part to the currents which the wires 

 are conveying. In the earlier forms of the apparatus, when 

 bare wire was used and the stand itself was rotated inside an 

 enclosure, the variation of the current through the galvano- 



