Ox THE MAGNETIC EFFECT OF ELECTRIC CONVECTION 131 



To calculate these currents we have two ways. Either we can con- 

 sider the electricity at rest and the motion of the disc through it to 

 produce an electromotive force in the direction of motion and propor- 

 tional to the velocity of motion, to the electrification, and to the surface 

 resistance; or, as Professor Helmholtz has suggested, we can consider 

 the electricity to move with the disc and as it comes to the edge of the 

 inductor to he set free to return by conduction currents to the other 

 edge of the inductor so as to supply the loss there. The problem is 

 capable of solution in the case of a disc without a hole in the centre but 

 the results are too complicated to be of much use. Hence scratches 

 were made on the disc in concentric circles about -6 cm. apart by which 

 the radial component of the currents was destroyed and the problem 

 became easily calculable. 



For, let the inductor cover -th part of the circumference of any 



n 



one of the conducting circles; then, if C is a constant, the current in 



the circle outside the inductor will be +-, and inside the area of the 



1 n 



inductor C^ n ~ l \ On the latter is superposed the convection cur- 



fi 



rent equal to -\-C. Hence the motion of electricity throughout the 



whole circle is - what it would have been had the inductor covered the 

 n 



whole circle. 



In one experiment n was about 8. By comparison with the other 

 experiments we know that had electric conduction alone produced effect 

 we should have observed at the telescope 5- mm. Had electric con- 

 vection alone produced magnetic effect we should have had -j- 5- 7 mm. 

 And if they both had effect it would have been -f- -7 mm., which is prac- 

 tically zero in the presence of so many disturbing causes. No effect 

 was discovered, or at least no certain effect, though every care was used. 

 Hence we may conclude with reasonable certainty that electricity pro- 

 duces nearly if not quite the same magnetic effect in the case of con- 

 vection as of conduction, provided the same quantity of electricity 

 passes a given point in the convection stream as in the conduction 

 stream. 



The currents in the disc were actually detected by using inductors 

 covering half the plate and placing the needle over the uncovered por- 

 tion; but the effect was too small to be measured accurately. To prove 



