Unit of Electrical Resistance. 163 



A metre long and square millimetre section has a symmetrical 

 sound; but, except in calculating the resistance of prisms of 

 mercury, 1 know of no practical advantage which it has over any 

 other magnitude. A metre of pure copper weighing a gramme 

 would be more practically useful, for it would serve as an imme- 

 diate term of comparison with the copper supplied, for conduct- 

 ing- wires. A mile of iron wire, of the size generally used in tele- 

 graphy, would have a practical advantage in finding the distance 

 of faults. I could give other definitions with practical advan- 

 tages, but do not know of any practical advantage which has re- 

 sulted, or could result, from a definition in terms of mercury. 



Then even in a mercury unit, no reason is assigned for prefer- 

 ring the particular definition given to any other, such as a metre 

 of mercury, weighing a gramme, or contained in a tube of say 

 1 millimetre diameter. I do not ask these questions as contend- 

 ing that these definitions are better than Dr. Siemens's, but 

 simply to point out the completely arbitrary nature of his defi- 

 nition. 



I have heard in favour of that definition two arguments which, 

 so far as they go, are really in its favour. First, that it was 

 coming into extensive use when the Standard Committee was 

 appointed ; secondly, that it was at least an intelligible definition. 

 The first argument is based on truth, though the use of Dr. Sie- 

 mens^ unit was by no means so general as some of his friends 

 suppose. No large English telegraph company used it, nor has 

 it, I believe, been at all adopted in France. Nevertheless I 

 admit that, owing to the undoubtedly excellent manufacture of 

 Messrs. Siemens' s resistance-coils, and the convenient arrange- 

 ments of those coils, his unit was largely used; but this ex- 

 tended use was not, I think, caused by the excellence of his de- 

 finition. People ordered coils from the most celebrated firm in 

 Europe and took what was given them — the miles of copper 

 wire before 1860, and the mercury units afterwards. 



This argument, however, was fully considered by the Com- 

 mittee, but was overthrown by the considerations that changes 

 were yearly being made in the coils supplied, that the definition, 

 being wholly arbitrary and having no intrinsic merit, could not 

 be compared in value with an approximation to the natural unit 

 adopted, and that the use was by no means so general as to 

 outweigh the two previous arguments. 



I next come to the question of intelligibility. No doubt most 

 people think they know what a metre of mercury of 1 millimetre 

 section means, and comparatively few understand the definition 

 adopted by the Committee. But who in practical life, or in the 

 use of standards, refers to their definition ? What Frenchman, 

 measuring the contents of a brick wall, thinks of the earth's 



M2 



