considered in connexion with the Dynamical Theory. 11 

 netism that a could give it, to current b, I obtained, 

 After 1 discharge, 21 per minute. 



In fact it was the general result of the experiments that 

 more magnetism is always given to the needle by arrangement 

 b than by «. In order, however, that the difference may be 

 striking, it is advisable not to approach too nearly the point of 

 magnetic saturation. The numbers quoted were obtained with 

 the condenser, which was still necessary, in order to make the 

 break sufficiently sudden. I have no doubt, however, that it 

 might have been dispensed with had the triple coil consisted 

 of a larger number of turns. 



The circumstances of this experiment are in some degree 

 represented by supposing, in the hydrodynamical analogue, one 

 of the balls to vary in size. When a given motion is suddenly 

 impressed on the other ball, the corresponding velocity gene- 

 rated in the first would vary inversely with its magnitude ; for 

 the larger the ball the greater hold, as it were, would it have 

 on the fluid. 



It is interesting also to examine the influence of neighbour- 

 ing soft iron on the character of the induced current. This in- 

 fluence is of two sorts ; but I refer here to the modifications 

 produced by the magnetic character of iron. The circulation 

 of induced currents in its mass may generally be prevented 

 from exercising any injurious influence on the result by using 

 only wires, or fragments of small size. The proximity of soft 

 iron always increases the coefficient of self-induction N, while 

 M may be either increased or diminished. The latter state- 

 ment is true also for the initial current y , which is propor- 

 tional to =r=. For the two wires of the coil A. however, it is 



easy to see that M and N are approximately equal, whether 

 there be soft iron in their neighbourhood or not. Thus, if A r 

 be connected with a Daniell cell while the circuit of A 2 is com- 

 pleted by the magnetizing spiral, the magnetism acquired by 

 the needle, after a break-induced current, is not much altered, 

 even if a considerable number of iron wires are placed in the 

 coil. The total current is increased fifteen times or more ; but 

 this is because the current lasts longer, the maximum or initial 

 value being no greater than before. This experiment strikingly 

 illustrates the comparative independence of the magnetizing 

 effect of a current on its duration. It seems probable a priori, 

 and is partly confirmed by some of my experiments, that this is 

 more especially true if we take the limiting magnetism which 



