Unit of Electrical Resistance. 167 



ing part of a coherent system for other electrical measure- 

 ments. 



Hitherto I have argued wholly on definitions, keeping in 

 reserve all the questions as to how far Dr. Siemens had been 

 successful in producing or reproducing his unit. For instance, 

 when above I say that the mercury unit is varying, I mean that 

 Dr. Siemens intends that it shall vary as better and better deter- 

 minations are made; and when I say that the B.A. unit is per- 

 manent, I mean that the Committee intend it to be permanent. 

 I will now consider how far some of the proposals for the ma- 

 nufacture have been, and are likely to be, practically successful. 

 In arguing this point I am quite willing to abandon all the dis- 

 crepancies hitherto pointed out in the units supplied by Dr. Sie- 

 mens. The important question undoubtedly is, not whether 

 Dr. Siemens has from first to last supplied faultless mercury 

 units, but whether any unit can be reproduced by a given length 

 and section or weight of mercury, or any other material. I think 

 I have shown that there is no reason to adopt a metre of mer- 

 cury of 1 millimetre section as unit of resistance ; but mercury 

 may nevertheless be the best substance of which to make units, 

 or by which to reproduce them. I am at a loss to conceive how 

 the two subjects can have been confused as they have been ; if pla- 

 tinum be the best metal for a standard weight, we are not there- 

 fore obliged to take a cubic metre of platinum as the unit weight. 

 It is undoubtedly of great importance to have some means of 

 reproducing a unit, in case the original be lost or altered ; and 

 Dr. Siemens's experiments are exceedingly valuable, although 

 they can hardly yet be said to be conclusive. Dr. Siemens be- 

 lieves that he can reproduce his unit (and therefore the B.A. 

 unit) at least to within 0*1 per cent, of accuracy, and, when the 

 greatest possible care is used, probably with an almost unlimited 

 exactness. It would, I am sure, give every member of the Com- 

 mittee great pleasure if this fact could be fully established • but 

 it is not yet established. I admit that the discrepancies between 

 the coils exhibited m 186.2 have little to do with this point; 

 although I was informed, as I now believe mistakenly, by Dr. 

 Esselbach, that the difference was due to a change in Dr. Sie- 

 mens^ standard. Mr. C. W. Siemens himself at the Royal 

 Society stated that inconvenience had been caused by a prema- 

 ture issue of coils based on an imperfect standard ; but after Dr. 

 Siemens's statements, I can feel no doubt that no material altera- 

 tion has knowingly occurred in this standard. Moreover I admit 

 that errors in early determinations would only show that the re- 

 production was not very easy, as no one now contends that it is. 

 I further allow that commercial coils cannot be accepted as 

 standards ; also that an error in the specific gravity of mercury 



