84 On the Self-induction of Wires. 



ferences from parallel wires, unless the intermediate apparatus 

 be double, one part being in circuit with one wire, the other 

 part (quite similar) in circuit with the other. In mentioning 

 my brother's system of bridge-working of telephones in 

 Part V., whereby the intermediate impedance is wholly re- 

 moved, I mentioned, without explanation, the cancelling of 

 inductive interferences. The present and preceding para- 

 graphs supply the needed explanation of that remark. The 

 intermediate apparatus, being in bridges across from one wire 

 to the other, does not in the least disturb the induction- 

 balance, so that transmission of speech is not interfered with 

 by foreign sounds. 



But theory goes much further than the above in predicting 

 interferences than practice up to the present time verifies. 

 For instance, if two perfectly equal wires be suspended at the 

 same height above the ground and be looped at the ends, ter- 

 minal telephones will not be interfered with by variations of 

 current in a parallel wire equidistant from both wires of the 

 loop-circuit, having its own circuit completed through the 

 earth. But if the loop-circuit be in a vertical plane, so that 

 one wire is at a greater height above the ground than the 

 other, there must be terminal disturbance produced, even 

 when the disturbing wire is equidistant. Similarly in the 

 many other cases of inequality that can be mentioned. 



The two matters, preservation of the induction-balance, and 

 transmission of signals in the same manner as on a single 

 wire, are intimately connected. If we have one, we also have 

 the other. The limitations of application of the method of 

 Part IV. may be summed up in saying that the loop -circuit 

 must either be far removed from all conductors, in which case 

 equivalence of the wires is quite needless ; or else they must 

 be equal in their electrical constants. In the latter case the 

 effective resistance R is the double of that of either wire, and 

 the effective capacity, inductance, and leakage are to be 

 measured as before described, whilst the variables are the 

 potential-difference of the wires and the current in each. But 

 the four electrical constants may vary in any (not too rapid) 

 manner along the line. And the impressed force (in the 

 investigations of Part IV.) may also be an arbitrary function 

 of the distance, provided it be put, half in one wire, half in the 

 other, oppositely directed in space. For, although equal, 

 similarly directed impressed forces will cause no terminal 

 disturbance (and none anywhere if other conductors be suffi- 

 ciently distant), yet disturbances at intermediate parts of the 

 line will result. It is true that the most practical case of 

 impressed force is when it is situated at one end only of the 



