364 Dr. Matthiessen on the so-called Mercury Unit. 



Again, the specific gravity of gold used is 19*27. The specific 

 gravity given in GmehVs ' Chemistry ' for gold wire is 19*3 to 

 19*4. Using the value 19'4, we find the conducting-power to be 

 36-72. The differences between the values obtained for the 

 mercury (namely, about 3 per cent.) show how difficult it is to 

 obtain good results with this metal when only ordinary care is 

 taken in experimenting. No doubt these differences are due to 

 the manner of calibrating and deducing the values for the diame- 

 ters of the tubes employed in my first experiments, published in 

 1858*; and when determining at a later period f the influence 

 of temperature on its conducting-power (finding with the one 

 tube we used nearly the same value), we did not think it neces- 

 sary to repeat the determination. The value given by Mr. 

 Hockin and myself is, as will presently be shown, probably the 

 correct one. 



In my papers on the conducting-power of metals and alloys, 

 I have advocated the use of the gold-silver alloy as a ready and 

 practical means, where only ordinary care is employed, of repro- 

 ducing resistances. Here, again, we find my statements correct, 

 for the values deduced by it agree well with the old ones, 

 especially the lead one, a metal which has been recommended by 

 Mr. Hockin and myself as eminently fit for the reproduction of 

 resistances where great care is employed J. Of course it is to be 

 hoped that no one will try to reproduce any unit by means of 

 this or any metal, now that the B. A. unit has been issued §; for 

 they will certainly not obtain such accurate measures (without an 

 enormous trouble and expense) as the copies of the B. A. unit at 

 present issued by the Committee appointed by the British Asso- 

 ciation; and as proof of the correctness of this statement, I would 

 draw the special attention of those who would endeavour to 

 reproduce a unit, first, to what has been stated in the Reports 

 of the Committee published in the Transactions of the British 

 Association, and secondly, to the remarks I am now about to make 

 on the sco-alled mercury unit. 



It is defined as the resistance of a column of mercury a metre 

 long and a square millimetre section at 0°. Now it may be 

 asserted : — 



1st. That no true mercury unit has been issued. 



2nd. That the units issued from time to time have not the same 

 resistance. 



1st. That no true mercury unit has been issued. This assertion 



* Phil. Trans. 1858, p. 383. t Ibid. 1862. 



X The specific gravity of lead is said not to alter on being hammered or 

 rolled (Reich, Journ. of Pract. Chem. vol. lxxviii. p. 730). 



§ Copies of which may be obtained from the Secretary to the Committee 

 (see Phil. Mag. vol. xxix. p. 248). 



