332 Messrs. Guthrio and Boys on 



case the form of the conductor is necessarily always the same 

 and its situation in the magnet is so too ; for its height is ad- 

 justed by the screw z till a horizontal line on the scale s is 

 seen on the cross-wires in the microscope. The strength of 

 the magnet is nearly constant ; and its variations are allowed 

 for by a method which will be described below : the speed of 

 the magnets can be accurately measured by the counter; and 

 therefore the torsion of the wire 10 gives an exact comparative 

 measure of the conductivity of the liquid in the vessel A, 

 provided that the torsion of the wire is uniform and that the 

 movement of the liquid in the cell has no influence on the 

 result. As the wire used cannot be made to carry 7 lbs., 

 but has in the course of the experiments to support the cell 

 together with a litre of liquid of any specific gravity between 

 those of water and oil of vitriol, its torsion is any thing but 

 constant from day to day; but its variations, due to these great 

 changes in the load and to slight changes in temperature, are 

 accurately allowed for at the same time that the change in the 

 strength of the magnet is taken into account ; and therefore, 

 if movement of liquid in the cell does not vitiate the results, 

 this method of comparing the conductivity of electrolytes is 

 free from error. 



To ascertain what effect movement of liquid in the cell has 

 on the result, it will be necessary to examine more closely 

 how the torsion of the wire is produced. When the magnets 

 are revolving, currents are induced in the liquid in a direction 

 to oppose the motion ; so the liquid is pushed round in the 

 direction in which the magnets are revolving by a force which 

 is directly proportional to the difference in speed between the 

 liquid and the magnets. If there were no friction of any kind 

 tending to resist the motion of the liquid, it would in time 

 attain a velocity equal to that of the magnets ; for so long- 

 as it was revolving more slowly than the magnets there 

 would be a force urging it on. But there is friction between 

 the outer layer of the liquid and the vessel, and between 

 each successive cylindrical layer and the one outside it; and 

 therefore the liquid never attains any great velocity at all. 

 But each elementary cylindrical layer soon reaches that speed 

 at which the force urging it forward is equal to the friction 

 between it and the layer outside ; and therefore the torsion 

 measured is exactly the same that would be obtained if it 

 were possible to integrate the force from the centre to the 

 edge of the vessel. If this reasoning is correct, the liquid 

 should be revolving fastest at the centre and more slowly 

 outside ; and this has been proved to be the case by a very 

 striking experiment. A solution of indigo was made of such 



