Electrical Resistances of Metallic Wires, 439 



Many theories have suggested themselves, but none are 

 entirely satisfactory. 



One theory was that, in coiling, the decrease of resistance 

 was due to a thickening of the wire on its internal side and a 

 thinning of the external side, thus bringing the centre of area 

 inwards, and so making the actual distance the current would 

 have to travel shorter, which would be equivalent to making 

 the wire shorter ; and then on uncoiling, the displacement of 

 the molecules, due to slipping on each other when seeking to 

 regain their original position, would have the effect of increa- 

 sing their distance apart, and consequently increasing the 

 resistance. But the weak point in this theory is that it will 

 not account for the change of signs which so often takes place, 

 nor will it account for the zinc wires giving increased resist- 

 ances on coiling. 



Another theory, which appears more promising, was that, 

 while being coiled and uncoiled, the wires were subject to 

 strains of a torsional kind, which might be due to their having 

 a certain amount of residual twist in them, owing to the way 

 in which they have been coiled on to the drums on the bench 

 of the wire drawer; and that on being coiled and uncoiled on 

 their bobbins this twist might be increased or decreased, so 

 giving different amounts in the resistance. But from experi- 

 ments I have made, which seem to me quite conclusive, it 

 appears that although there is a certain amount of twisting 

 and untwisting produced by being wound on to the bobbins, 

 yet this fails to account for the residue of the observed changes. 



With reference to the effects on the permanency of value of 

 a set of resistance-coils which are only subjected to one wind- 

 ing, I have not yet sufficient data to lay before you ; but, so 

 far as I have gone into this part of the subject, I believe that 

 the immediate alterations on coiling the w T ires on the bobbins 

 are to a great extent affected by the tension and the speed 

 with which they are wound, the thickness and kind of insu- 

 lator employed (cotton-covered wires appear to be less affected 

 than silk-covered wires), and also by the diameter of the bob- 

 bins on which the wire is wound (the smaller the bobbin the 

 greater the alteration) . But further experiments are necessary 

 to put these beyond doubt. 



Appended to this paper are a few tables showing the actual 

 readings in the first series of several of the wires tested (these 

 readings include the resistance of the leading wires, '0274). 

 These, together with the curves (PI. IX.), may be of some 

 service in assisting to a conclusion as to the probable cause of 

 the changes in the resistance of wires produced by coiling and 

 uncoiling. 



