Effects of Mechanical Stress on Electrical Resistance. 289 



section of the wire must be measured, and this is a very difficult 

 thing to do, even in the case of moderately thick wires, when 

 accuracy is required, since an error in the measurement of the 

 diameter is more than doubled in the value of the section. Everyone 

 who has tried to measure accurately the diameters of wires is well 

 aware of the great difficulty in doing so. Even although the one 

 measurement were quite accurate, the diameter at the place measured 

 would, in all probability, not be the average diameter of the wire. 

 This can, of course, be corrected by taking measurements at a 

 number of places, and taking the average, but the process is very 

 tedious, and, when done, is not thoroughly trustworthy, for it is 

 so easy to make errors in using the ordinary micrometer gauge. If 

 the section be determined by the longer method from a measurement 

 of the density, it is, of course, more accurately obtained, but, even 

 then, the weighing in water of small lengths of wire is an uncertain 

 thing ; whereas, for the length per unit of weight, the weighing has 

 only to be made in air, and therefore the error due to weighing in 

 water is avoided. Clearly, then, the "weight specific resistance" 

 gives much more accurate results for the total resistance of a wire 

 than the "volume specific resistance," particularly in the case of 

 practical work, where very little care is taken in measuring the 

 diameter. Certainly, it would have been incomparably more difficult, 

 and would have taken a much longer time to obtain the results given 

 in this paper, had the " volume specific resistance " been used. As 

 will be shown further on, it is not necessary, for mere comparison of 

 two specific resistances, even to measure the length per unit of 

 weight. 



Dr. A. Matthiessen, F.R.S., in his paper " On the Specific Resist- 

 ance of Metals in terms of the British Association Unit (1864) of 

 Electric Resistance, &c." (' Phil. Mag.,' May, 1865), says : 



" We employed the length weight in preference to the length 

 section, knowing that the weight of a wire may be much more accu- 

 rately determined than its section, whether deduced directly from the 

 measurements of the diameter, or indirectly from the specific gravity, 

 the determination of the latter introducing an error. Of course, in 

 endeavouring to reproduce resistances, it is wise to avoid the use of 

 unnecessary values ; and it is just as well, and certainly a much more 

 accurate method, to determine a resistance in length weight than in 

 length section." 



The term " specific resistance " was introduced by Weber, but by 

 it he meant the weight unit, and always used it unless when otherwise 

 stated. Lord Kelvin points this out in his paper on "Measurement 

 of Electromotive Forces in Absolute Measure " (vol. 1, ' Math, and 

 Phys. Papers '), and throughout three papers relating to the present 

 investigation, on " Electromagnetic Qualities of Metals," " Analytical 



