126 Dr. 0. J. Lodge on Intermittent Currents 



An obvious modification is, to arrange both speaking and 

 bearing telephone in circuit with both the parallel coiled 

 wires in series, first, so that the currents flow the same way 

 in both wires, and then with one wire reversed ; so that in 

 the first case the self-induction is a maximum, and in the other 

 almost zero. 



Any one who tries this experiment will at once appreciate 

 the very decided difference in the telephonic indications. In 

 fact, resistance is of much less consequence in a telephone- 

 circuit than is self-induction. This fact obviously explains 

 the discrepancy which has been observed between experiments 

 on resistance-bobbins thrown into a telephonic circuit in a 

 laboratory, and the same resistance when forming an actual 

 line. The resistance-coils are wound so as to have no self- 

 induction ; and they accordingly weaken the effect very little. 

 It is surprising how many thousand ohms can be thrown into 

 a talking circuit without stopping a conversation. But if the 

 bobbins were wound all in one direction, they would be still 

 worse than a line-resistance. 



3. It is plain from this that the only right method of arran- 

 ging a telephone-line is to have two wires running close 

 together — one for the direct, the other for the return current. 

 This arrangement has, I believe, been used in order to avoid 

 induction-effects from battery-currents in neighbouring wires; 

 but it has also the great advantage of diminishing the self- 

 induction of the wire itself. It is, however, a question how 

 far the increase in electrostatic capacity would militate against 

 this advantage in a very long land-line. It is certain to be 

 injurious as tar as it goes; but if the wires are very thin, their 

 electrostatic capacity will be but small. 



In the case of a cable the double wire would have decidedly 

 less capacity than a single wire ; and the capacity diminishes 

 rapidly with the size of the wire. Hence it appears that a 

 telephone-cable should have a pair of exceedingly fine insu- 

 lated wires of the very highest conductivity pretty close toge- 

 ther in a large insulating sheath. 



I need hardly say, perhaps, that no advantage will be gained 

 by attempting to do away with the self-induction of the tele- 

 phone-bobbin itself by means of a second wire wound along- 

 side the first and forming a closed circuit — because the inter- 

 mittent currents induced in this would be nearly as strong as 

 those in the direct circuit, and, being always in an opposite di- 

 rection, will nearly neutralize their intermittent effect on 

 external coils or on the telephone-plate. 



I propose now to work out the theory of the induction- 

 balance, and of one or two other cases of intermittent currents, 



