intended to reduce the effects of Inductive Action. 339 



discharges from the dielectric, since any flow occasioned by them 

 which could possibly pass through that point would necessarily 

 be from positive to negative, or in the direction of the first 

 current, slightly prolonging its action. The most important 

 question, however, connected with this arrangement, and which 

 can only be decided by experiment, is, whether the connexion 

 of the two oppositely electrified insulated remote ends with an 

 insulated galvanometer would not produce an equal effect with 

 a smaller amount of current, since its velocity would be doubled. 

 Let us consider an analogous example. 



Take two Leyden jars precisely equal in capacity, that is to 

 say, which will, with a given number of turns of the electrical 

 machine, discharge through precisely the same interval and 

 produce the same heating effect in the thermo-electrometer. 

 Connect these jars respectively with the positive and negative 

 conductors of the machine, so as to charge one positively and 

 the other negatively. Connect with each jar a discharging 

 electrometer, arranged so as to discharge each to its own coating 

 through a given interval, and count the number of turns re- 

 quired for the effect. The two jars will discharge simultane- 

 ously, or very nearly so. 



Now remove the discharging electrometers, connect the outer 

 coatings of the jars with each other by a conductor, and connect 

 a discharging electrometer with the two inner coatings, so that 

 they may discharge through it to each other, the interval being 

 made the same as before. The discharge will now pass with 

 half the number of turns, the respective tensions of each jar 

 mutually operating to produce the effect. 



Again, make the discharging interval between the jars double ; 

 and the discharge will take place with the same number of 

 turns as in the first case, but the spark will be twice as long. 

 We here observe, then, that the effect of the mutual attraction 

 of the positively and negatively electrified conductors is to faci- 

 litate the flow of electricity between them ; and it is just upon 

 this principle that I have been led to the conclusion that an 

 insulated return wire would be preferable to an earth-circuit. 

 Induction between the two cables, it is true, cannot be pre- 

 vented ; but it will be much less, as I have already shown, with 

 two separate cables than with two wires included in the same 

 sheath. I have been thus explicit, because the tendency of some 

 scientific discussions which have lately taken place has been 

 rather unfavourable to the recognition of this principle; and 

 whilst I would not dogmatically enforce my views, yet I think 

 they would bear the test of trial. 



The success of the foregoing ])lan, however, would be, after 

 all, only palliating the evil vvitliout removing or mitigating the 



