Prof. Norton on Molecular Physics. 287 



arc afc becomes negative in equation (b). Equations (a) and 

 (c) fail for y=0. 



The investigation here made proceeds upon the supposition 

 that the breadth of the magnet is indefinitely small. If we sup- 

 pose it to be indefinitely great, the action of each individual 

 transverse current upon any point / (fig. 10) would be inversely 

 proportional to the distance of the current from this point *, and 

 it will be readily seen that the amount of force propagated to /, 

 within any angle, as mf r, will be the same whatever may be 

 the value offn. 



The equation for the value of the effective impulsive force will 

 be approximately of the form 



w=k(-+ -A 

 \n n'J 



k being a constant coefficient dependent upon the strength of 

 the magnet, / and V the parts of the length a b of the magnet 

 comprised between the angles afc and bfd subtended by the 

 two ends, and n n 1 the mean distances of these parts from /. 

 We approximate to this state of things in proportion as the 

 magnet is supposed to be broader, and shorter and thinner, or 

 in proportion as, with a magnet of given dimensions, the point 

 yis taken nearer to the magnet. 



Let us now replace Faraday's lines of force by the curves of 

 equal impulsive force of the magnet, and consider what should 

 be the effect of moving a wire across them, along any line m r, 

 ms, mt y &c. (fig. 9)/ 



It is obvious that if the movement be outward, the impulsive 

 force taking effect upon the wire will decrease ; and that if it be 

 inward, the force will increase. Hence, agreeably to the funda- 

 mental principle before alluded to (p. 283), in the first case there 

 should be an induced current having the same direction as the 

 currents of the upper face of the magnet, and in the second 

 case a current pursuing the opposite direction. Again, the 

 amount of change of force which results from the displacement 

 of the wire, and therefore the quantity of electricity which 

 this change sets in motion, should depend solely upon the 

 number of curves traversed. We may add that in whatever 

 part of the magnetic field, and in whatever direction the wire be 

 supposed to move, the theoretical result is in perfect accordance 

 with the facts as experimentally established by Faraday. 



In the foregoing we have supposed the wire, transverse to the 

 magnet, to be moved parallel to itself to various points of the 



* The individual molecular currents lying in a transverse section of the 

 magnet are here supposed to be replaced by two linear currents transverse 

 to the magnet, one in the upper and the other in the lower surface. 



