ON THE ABSOLUTE UNIT OF ELECTRICAL EESISTANCE 153 



One of the principal criticisms I have to offer with respect to this 

 method is the great numher of quantities difficult to observe, which 

 enter the equation as squares, cubes, or even fourth powers. Thus S 2 

 depends upon the fourth power of the radius of the earth inductor. 

 Now this earth inductor was wound years before by W. Weber, and the 

 mean radius determined from the length of wire and controlled by 

 measuring the circumference of the layers. Now the wire was nearly 

 3-2 mm. diameter with its coating, and the outer and inner radii were 

 115- mm. and 142 mm. Hence the diameter of the wire occupied two 

 per cent of the radius of the coil, making it uncertain to what point 

 the radius should be measured. As the coil is wound, each winding 

 sinks into the space between the two wires beneath, except at one spot 

 where it must pass over the tops of the lower wires. The wire must 

 also be wound in a helix. All these facts tend to diminish 8 and make 

 its value as deduced from the length of the wire too large; and any 

 kinks or irregularities in the wire tend in the same direction. And 

 these errors must be large in an earth-inductor of such dimensions, 

 where the wire is so large and many layers are piled on each other. 

 If we admit an error of one-half a millimetre in the radius as deter- 

 mined in this way, it would diminish the value of S 2 1-4 per cent, and 

 make Kohlrausch's result only -6 per cent greater than the result of 

 the British Association Committee. 



Three other quantities, T, X and K, are very hard to determine with 

 accuracy, and yet T enters as a square. It is to be noted that this 

 earth-inductor is the same as that used by Weber in his experiment of 

 1862, and which also gave a larger value to the Ohm than those of the 

 British Association Committee. Indeed, the results with this inductor 

 and by this method form the only cases where the absolute resistance of the 

 Ohm has been found greater than that from the experiments of the British 

 Association Committee, 



There seems to be a small one-sided error in A and B which Kohl- 

 rausch does not mention, but which Weber, in his old experiments of 

 1851, considered worthy of a -6 per cent correction, and which would 



diminish by 1-2 per cent. This is the error due to loss of 



time in turning the earth-inductor. As Kohlrausch's needle had a 

 longer time of vibration than Weber's, the correction will be much 

 smaller. In Weber's estimate the damping was not taken into account, 

 and indeed it is impossible to do so with exactness. To get some idea 

 of the value of the correction, however, we can assume that the current 



